CA2_MLP/NN_HW2_MohsenFayyaz.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "id": "hbTFekX_9GCV",
    "pycharm": {
     "is_executing": false
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Mohsen\\Anaconda3\\lib\\site-packages\\statsmodels\\tools\\_testing.py:19: FutureWarning: pandas.util.testing is deprecated. Use the functions in the public API at pandas.testing instead.\n",
      "  import pandas.util.testing as tm\n"
     ]
    }
   ],
   "source": [
    "# ! nvidia-smi\n",
    "from keras.datasets import cifar10\n",
    "import tensorflow as tf\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import logging\n",
    "from sklearn.metrics import classification_report, confusion_matrix\n",
    "from time import time\n",
    "import seaborn as sns\n",
    "import pandas as pd\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "tf.autograph.set_verbosity(0, False)\n",
    "logging.getLogger(\"tensorflow\").setLevel(logging.ERROR)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "XBNR5WpbcAVo"
   },
   "source": [
    "# Q1 - MLP (Classification)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "GZ9KeDVOg383"
   },
   "source": [
    "## Download & Visualize"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "DgdOWm9fgP_b"
   },
   "outputs": [],
   "source": [
    "(x_train, y_train), (x_test, y_test) = cifar10.load_data()\n",
    "print(\"x_train:\", x_train.shape)\n",
    "print(\"y_train:\", y_train.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "HZzCknZbgZVr"
   },
   "outputs": [],
   "source": [
    "def plot_pics(x: np.ndarray, y: np.ndarray):\n",
    "    fig = plt.figure(figsize=(15, 4))\n",
    "    for class_num in range(10):\n",
    "        class_indices = np.where(y == class_num)[0]\n",
    "        class_x = x[class_indices]\n",
    "        rnd_idx = np.random.randint(len(class_x))\n",
    "        image = class_x[rnd_idx]\n",
    "        ax = fig.add_subplot(1, 10, class_num + 1, xticks=[], yticks=[])\n",
    "        ax.set_title(class_num)\n",
    "        plt.imshow(image)\n",
    "    plt.show()\n",
    "\n",
    "for i in range(3):\n",
    "    plot_pics(x_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "AxTJlfj_Dusv"
   },
   "source": [
    "## Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "1J1I4CtpiK9M"
   },
   "outputs": [],
   "source": [
    "def onehot(a):\n",
    "    a = a.flatten()\n",
    "    o = np.zeros((a.size, a.max() + 1))\n",
    "    o[np.arange(a.size), a] = 1\n",
    "    return o\n",
    "\n",
    "def preprocess(x, y):\n",
    "    assert x.shape[1:] == (32, 32, 3)\n",
    "    assert y.shape[1:] == (1, )\n",
    "    x = np.mean(x, axis=-1)  # grayscale\n",
    "    x = np.reshape(x, (-1, 32 * 32))  # flatten\n",
    "    x /= 255  # normalize\n",
    "    y = onehot(y)  # onehot labels\n",
    "    return x, y\n",
    "\n",
    "x_train_processed, y_train_processed = preprocess(x_train, y_train)\n",
    "x_test_processed, y_test_processed = preprocess(x_test, y_test)\n",
    "x_train_processed.shape, y_train_processed.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "buLmvfv6K7Zo"
   },
   "source": [
    "## Model & Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {
    "id": "U1cMk89jG8zM",
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "class Trainer:\n",
    "    def __init__(self, units=1024, activation_function=\"relu\", \n",
    "                 optimizer=\"adam\", n_layers=4, loss_function='categorical_crossentropy', input_size=1024) -> None:\n",
    "        self.optimizer = optimizer\n",
    "        self.units = units\n",
    "        self.activation_function = activation_function\n",
    "        self.n_layers = n_layers\n",
    "        self.loss_function = loss_function\n",
    "        self.input_size = input_size\n",
    "        self.model = self.build_mlp_model()\n",
    "        self.history = None\n",
    "        self.training_time = None\n",
    "        \n",
    "    \n",
    "    def print_summary(self):\n",
    "        print(\"######################\")\n",
    "        print(f\"Training Time: {self.training_time:.2f}s\")\n",
    "        print(\"batch size:\", self.batch_size)\n",
    "        print(\"n_layers:\", self.n_layers)\n",
    "        print(\"optimizer:\", self.optimizer)\n",
    "        print(\"activation_function:\", self.activation_function)\n",
    "        print(\"loss_function\", self.loss_function)\n",
    "        print(self.model.summary())\n",
    "\n",
    "    def build_mlp_model(self):\n",
    "        model = tf.keras.Sequential([tf.keras.layers.Input(shape=(self.input_size, ))])\n",
    "        for n in range(self.n_layers):\n",
    "            model.add(tf.keras.layers.Dense(self.units / 2**n, activation=self.activation_function))\n",
    "            model.add(tf.keras.layers.Dropout(0.1))\n",
    "        model.add(tf.keras.layers.Dense(10, activation='softmax'))\n",
    "        \n",
    "        model.compile(\n",
    "            optimizer=self.optimizer,\n",
    "            loss=self.loss_function,\n",
    "            metrics=['accuracy']\n",
    "        )\n",
    "        return model\n",
    "    \n",
    "    def train(self, x_train, y_train, batch_size=32, epochs=50):\n",
    "        self.batch_size = batch_size\n",
    "        assert x_train.shape[1:] == (self.input_size, )\n",
    "        assert y_train.shape[1:] == (10, )\n",
    "        early_stopping = tf.keras.callbacks.EarlyStopping(\n",
    "            monitor='val_loss', \n",
    "            verbose=1,\n",
    "            patience=10,\n",
    "            mode='min',\n",
    "            restore_best_weights=True\n",
    "        )\n",
    "        start = time()\n",
    "        self.history = self.model.fit(\n",
    "            x_train, y_train,\n",
    "            batch_size=batch_size,\n",
    "            epochs=epochs,\n",
    "            verbose=1,\n",
    "            validation_split=0.2,\n",
    "            callbacks=[early_stopping],\n",
    "        )\n",
    "        self.training_time = time() - start\n",
    "    \n",
    "    def plot_history(self):\n",
    "        fig = plt.figure(figsize=(12, 4))\n",
    "        metrics = ['loss', 'accuracy']\n",
    "        for n, metric in enumerate(metrics):\n",
    "            plt.subplot(1, 2, n+1)\n",
    "            plt.plot(self.history.epoch, self.history.history[metric], label='Train')\n",
    "            plt.plot(self.history.epoch, self.history.history[f\"val_{metric}\"], linestyle=\"--\", label='Validation')\n",
    "            plt.xlabel('Epoch')\n",
    "            plt.ylabel(metric)\n",
    "            plt.title(metric)\n",
    "        plt.legend()\n",
    "        plt.show()\n",
    "\n",
    "    \n",
    "    def evaluate(self, x_test, y_test):\n",
    "        assert x_test.shape[1:] == (self.input_size, )\n",
    "        assert y_test.shape[1:] == (10, )\n",
    "        [test_loss,test_acc] = self.model.evaluate(x_test,y_test)\n",
    "        print(\"Test Loss:\", test_loss, \"Test Accuracy:\", test_acc)\n",
    "        test_preds = np.argmax(self.model.predict(x_test), axis=-1)\n",
    "        y_test = np.argmax(y_test, axis=-1)\n",
    "        print(classification_report(y_test, test_preds))\n",
    "        self.plot_cm(y_test, test_preds)\n",
    "    \n",
    "    def plot_cm(self, y_true, preds):\n",
    "        cm = confusion_matrix(y_true, preds)\n",
    "        plt.figure(figsize=(7, 5))\n",
    "        ax = sns.heatmap(cm, annot=True, fmt=\"d\")\n",
    "        bottom, top = ax.get_ylim()\n",
    "        ax.set_ylim(bottom + 0.5, top - 0.5)\n",
    "        plt.title('Confusion Matrix')\n",
    "        plt.ylabel('True Label')\n",
    "        plt.xlabel('Predicted Label')\n",
    "        plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "Xu4X5eod6UuZ"
   },
   "outputs": [],
   "source": [
    "for n_units in [128, 1024, 2048]:\n",
    "    trainer = Trainer(\n",
    "        units=n_units,\n",
    "        activation_function=\"relu\", \n",
    "        optimizer=tf.keras.optimizers.Adam(learning_rate=1e-4), \n",
    "        n_layers=4\n",
    "    )\n",
    "    trainer.train(x_train_processed, y_train_processed, \n",
    "                  batch_size=64, \n",
    "                  epochs=100)\n",
    "    trainer.print_summary()\n",
    "    trainer.plot_history()\n",
    "    trainer.evaluate(x_test_processed, y_test_processed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "for batch_size in [32, 64, 128, 512]:\n",
    "    trainer = Trainer(\n",
    "        units=1024,\n",
    "        activation_function=\"relu\", \n",
    "        optimizer=tf.keras.optimizers.Adam(learning_rate=1e-4), \n",
    "        n_layers=4\n",
    "    )\n",
    "    trainer.train(x_train_processed, y_train_processed, \n",
    "                  batch_size=batch_size, \n",
    "                  epochs=200)\n",
    "    trainer.print_summary()\n",
    "    trainer.plot_history()\n",
    "    trainer.evaluate(x_test_processed, y_test_processed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "for activation_function in [\"softmax\", \"sigmoid\", \"tanh\", \"relu\"]:\n",
    "    trainer = Trainer(\n",
    "        units=1024,\n",
    "        activation_function=activation_function, \n",
    "        optimizer=tf.keras.optimizers.Adam(learning_rate=2e-4), \n",
    "        n_layers=4\n",
    "    )\n",
    "    trainer.train(x_train_processed, y_train_processed, \n",
    "                  batch_size=128, \n",
    "                  epochs=500)\n",
    "    trainer.print_summary()\n",
    "    trainer.plot_history()\n",
    "    trainer.evaluate(x_test_processed, y_test_processed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "for loss_function in [\"cosine_similarity\", \"kl_divergence\", \"mean_squared_error\", \"categorical_crossentropy\"]:\n",
    "    trainer = Trainer(\n",
    "        units=1024,\n",
    "        activation_function=\"relu\", \n",
    "        optimizer=tf.keras.optimizers.Adam(learning_rate=2e-4),\n",
    "        n_layers=4,\n",
    "        loss_function = loss_function\n",
    "    )\n",
    "    trainer.train(x_train_processed, y_train_processed, \n",
    "                  batch_size=128, \n",
    "                  epochs=500)\n",
    "    trainer.print_summary()\n",
    "    trainer.plot_history()\n",
    "    trainer.evaluate(x_test_processed, y_test_processed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "optimizers = [\n",
    "    tf.keras.optimizers.SGD(learning_rate=1e-2),\n",
    "    tf.keras.optimizers.RMSprop(learning_rate=2e-4),\n",
    "    tf.keras.optimizers.Adam(learning_rate=2e-4)\n",
    "]\n",
    "for optimizer in optimizers:\n",
    "    trainer = Trainer(\n",
    "        units=1024,\n",
    "        activation_function=\"relu\", \n",
    "        optimizer=optimizer, \n",
    "        n_layers=4,\n",
    "        loss_function = \"kl_divergence\"\n",
    "    )\n",
    "    trainer.train(x_train_processed, y_train_processed, \n",
    "                  batch_size=128, \n",
    "                  epochs=500)\n",
    "    trainer.print_summary()\n",
    "    trainer.plot_history()\n",
    "    trainer.evaluate(x_test_processed, y_test_processed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "for n_layers in [3, 4, 5]:\n",
    "    trainer = Trainer(\n",
    "        units=1024,\n",
    "        activation_function=\"relu\", \n",
    "        optimizer=tf.keras.optimizers.Adam(learning_rate=2e-4), \n",
    "        n_layers=n_layers,\n",
    "        loss_function = \"kl_divergence\"\n",
    "    )\n",
    "    trainer.train(x_train_processed, y_train_processed, \n",
    "                  batch_size=128, \n",
    "                  epochs=500)\n",
    "    trainer.print_summary()\n",
    "    trainer.plot_history()\n",
    "    trainer.evaluate(x_test_processed, y_test_processed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "### BEST\n",
    "trainer = Trainer(\n",
    "    units=1024,\n",
    "    activation_function=\"relu\", \n",
    "    optimizer=tf.keras.optimizers.Adam(learning_rate=2e-4), \n",
    "    n_layers=4,\n",
    "    loss_function = \"kl_divergence\"\n",
    ")\n",
    "trainer.train(x_train_processed, y_train_processed, \n",
    "              batch_size=128, \n",
    "              epochs=500)\n",
    "trainer.print_summary()\n",
    "trainer.plot_history()\n",
    "trainer.evaluate(x_test_processed, y_test_processed)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Q2 - MLP (Regression)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(date, 70), (price, 1741), (bedrooms, 10), (bathrooms, 26), (sqft_living, 566), (sqft_lot, 3113), (floors, 6), (waterfront, 2), (view, 5), (condition, 5), (sqft_above, 511), (sqft_basement, 207), (yr_built, 115), (yr_renovated, 60), (street, 4525), (city, 44), (statezip, 77), (country, 1), "
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>date</th>\n",
       "      <th>price</th>\n",
       "      <th>bedrooms</th>\n",
       "      <th>bathrooms</th>\n",
       "      <th>sqft_living</th>\n",
       "      <th>sqft_lot</th>\n",
       "      <th>floors</th>\n",
       "      <th>waterfront</th>\n",
       "      <th>view</th>\n",
       "      <th>condition</th>\n",
       "      <th>sqft_above</th>\n",
       "      <th>sqft_basement</th>\n",
       "      <th>yr_built</th>\n",
       "      <th>yr_renovated</th>\n",
       "      <th>street</th>\n",
       "      <th>city</th>\n",
       "      <th>statezip</th>\n",
       "      <th>country</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2014-05-02 00:00:00</td>\n",
       "      <td>3.130000e+05</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.50</td>\n",
       "      <td>1340</td>\n",
       "      <td>7912</td>\n",
       "      <td>1.5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>1340</td>\n",
       "      <td>0</td>\n",
       "      <td>1955</td>\n",
       "      <td>2005</td>\n",
       "      <td>18810 Densmore Ave N</td>\n",
       "      <td>Shoreline</td>\n",
       "      <td>WA 98133</td>\n",
       "      <td>USA</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2014-05-02 00:00:00</td>\n",
       "      <td>2.384000e+06</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.50</td>\n",
       "      <td>3650</td>\n",
       "      <td>9050</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>3370</td>\n",
       "      <td>280</td>\n",
       "      <td>1921</td>\n",
       "      <td>0</td>\n",
       "      <td>709 W Blaine St</td>\n",
       "      <td>Seattle</td>\n",
       "      <td>WA 98119</td>\n",
       "      <td>USA</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2014-05-02 00:00:00</td>\n",
       "      <td>3.420000e+05</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.00</td>\n",
       "      <td>1930</td>\n",
       "      <td>11947</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>1930</td>\n",
       "      <td>0</td>\n",
       "      <td>1966</td>\n",
       "      <td>0</td>\n",
       "      <td>26206-26214 143rd Ave SE</td>\n",
       "      <td>Kent</td>\n",
       "      <td>WA 98042</td>\n",
       "      <td>USA</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2014-05-02 00:00:00</td>\n",
       "      <td>4.200000e+05</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.25</td>\n",
       "      <td>2000</td>\n",
       "      <td>8030</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>1000</td>\n",
       "      <td>1000</td>\n",
       "      <td>1963</td>\n",
       "      <td>0</td>\n",
       "      <td>857 170th Pl NE</td>\n",
       "      <td>Bellevue</td>\n",
       "      <td>WA 98008</td>\n",
       "      <td>USA</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2014-05-02 00:00:00</td>\n",
       "      <td>5.500000e+05</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.50</td>\n",
       "      <td>1940</td>\n",
       "      <td>10500</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>1140</td>\n",
       "      <td>800</td>\n",
       "      <td>1976</td>\n",
       "      <td>1992</td>\n",
       "      <td>9105 170th Ave NE</td>\n",
       "      <td>Redmond</td>\n",
       "      <td>WA 98052</td>\n",
       "      <td>USA</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4595</th>\n",
       "      <td>2014-07-09 00:00:00</td>\n",
       "      <td>3.081667e+05</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.75</td>\n",
       "      <td>1510</td>\n",
       "      <td>6360</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>1510</td>\n",
       "      <td>0</td>\n",
       "      <td>1954</td>\n",
       "      <td>1979</td>\n",
       "      <td>501 N 143rd St</td>\n",
       "      <td>Seattle</td>\n",
       "      <td>WA 98133</td>\n",
       "      <td>USA</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4596</th>\n",
       "      <td>2014-07-09 00:00:00</td>\n",
       "      <td>5.343333e+05</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.50</td>\n",
       "      <td>1460</td>\n",
       "      <td>7573</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>1460</td>\n",
       "      <td>0</td>\n",
       "      <td>1983</td>\n",
       "      <td>2009</td>\n",
       "      <td>14855 SE 10th Pl</td>\n",
       "      <td>Bellevue</td>\n",
       "      <td>WA 98007</td>\n",
       "      <td>USA</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4597</th>\n",
       "      <td>2014-07-09 00:00:00</td>\n",
       "      <td>4.169042e+05</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.50</td>\n",
       "      <td>3010</td>\n",
       "      <td>7014</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>3010</td>\n",
       "      <td>0</td>\n",
       "      <td>2009</td>\n",
       "      <td>0</td>\n",
       "      <td>759 Ilwaco Pl NE</td>\n",
       "      <td>Renton</td>\n",
       "      <td>WA 98059</td>\n",
       "      <td>USA</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4598</th>\n",
       "      <td>2014-07-10 00:00:00</td>\n",
       "      <td>2.034000e+05</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.00</td>\n",
       "      <td>2090</td>\n",
       "      <td>6630</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>1070</td>\n",
       "      <td>1020</td>\n",
       "      <td>1974</td>\n",
       "      <td>0</td>\n",
       "      <td>5148 S Creston St</td>\n",
       "      <td>Seattle</td>\n",
       "      <td>WA 98178</td>\n",
       "      <td>USA</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4599</th>\n",
       "      <td>2014-07-10 00:00:00</td>\n",
       "      <td>2.206000e+05</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.50</td>\n",
       "      <td>1490</td>\n",
       "      <td>8102</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>1490</td>\n",
       "      <td>0</td>\n",
       "      <td>1990</td>\n",
       "      <td>0</td>\n",
       "      <td>18717 SE 258th St</td>\n",
       "      <td>Covington</td>\n",
       "      <td>WA 98042</td>\n",
       "      <td>USA</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4600 rows × 18 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                     date         price  bedrooms  bathrooms  sqft_living  \\\n",
       "0     2014-05-02 00:00:00  3.130000e+05       3.0       1.50         1340   \n",
       "1     2014-05-02 00:00:00  2.384000e+06       5.0       2.50         3650   \n",
       "2     2014-05-02 00:00:00  3.420000e+05       3.0       2.00         1930   \n",
       "3     2014-05-02 00:00:00  4.200000e+05       3.0       2.25         2000   \n",
       "4     2014-05-02 00:00:00  5.500000e+05       4.0       2.50         1940   \n",
       "...                   ...           ...       ...        ...          ...   \n",
       "4595  2014-07-09 00:00:00  3.081667e+05       3.0       1.75         1510   \n",
       "4596  2014-07-09 00:00:00  5.343333e+05       3.0       2.50         1460   \n",
       "4597  2014-07-09 00:00:00  4.169042e+05       3.0       2.50         3010   \n",
       "4598  2014-07-10 00:00:00  2.034000e+05       4.0       2.00         2090   \n",
       "4599  2014-07-10 00:00:00  2.206000e+05       3.0       2.50         1490   \n",
       "\n",
       "      sqft_lot  floors  waterfront  view  condition  sqft_above  \\\n",
       "0         7912     1.5           0     0          3        1340   \n",
       "1         9050     2.0           0     4          5        3370   \n",
       "2        11947     1.0           0     0          4        1930   \n",
       "3         8030     1.0           0     0          4        1000   \n",
       "4        10500     1.0           0     0          4        1140   \n",
       "...        ...     ...         ...   ...        ...         ...   \n",
       "4595      6360     1.0           0     0          4        1510   \n",
       "4596      7573     2.0           0     0          3        1460   \n",
       "4597      7014     2.0           0     0          3        3010   \n",
       "4598      6630     1.0           0     0          3        1070   \n",
       "4599      8102     2.0           0     0          4        1490   \n",
       "\n",
       "      sqft_basement  yr_built  yr_renovated                    street  \\\n",
       "0                 0      1955          2005      18810 Densmore Ave N   \n",
       "1               280      1921             0           709 W Blaine St   \n",
       "2                 0      1966             0  26206-26214 143rd Ave SE   \n",
       "3              1000      1963             0           857 170th Pl NE   \n",
       "4               800      1976          1992         9105 170th Ave NE   \n",
       "...             ...       ...           ...                       ...   \n",
       "4595              0      1954          1979            501 N 143rd St   \n",
       "4596              0      1983          2009          14855 SE 10th Pl   \n",
       "4597              0      2009             0          759 Ilwaco Pl NE   \n",
       "4598           1020      1974             0         5148 S Creston St   \n",
       "4599              0      1990             0         18717 SE 258th St   \n",
       "\n",
       "           city  statezip country  \n",
       "0     Shoreline  WA 98133     USA  \n",
       "1       Seattle  WA 98119     USA  \n",
       "2          Kent  WA 98042     USA  \n",
       "3      Bellevue  WA 98008     USA  \n",
       "4       Redmond  WA 98052     USA  \n",
       "...         ...       ...     ...  \n",
       "4595    Seattle  WA 98133     USA  \n",
       "4596   Bellevue  WA 98007     USA  \n",
       "4597     Renton  WA 98059     USA  \n",
       "4598    Seattle  WA 98178     USA  \n",
       "4599  Covington  WA 98042     USA  \n",
       "\n",
       "[4600 rows x 18 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "raw_df = pd.read_csv('data.csv')\n",
    "for col in raw_df.columns:\n",
    "    print(f\"({col}, {len(raw_df[col].unique())})\", end=\", \")\n",
    "raw_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## a)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>date</th>\n",
       "      <th>price</th>\n",
       "      <th>bedrooms</th>\n",
       "      <th>bathrooms</th>\n",
       "      <th>sqft_living</th>\n",
       "      <th>sqft_lot</th>\n",
       "      <th>floors</th>\n",
       "      <th>waterfront</th>\n",
       "      <th>view</th>\n",
       "      <th>sqft_above</th>\n",
       "      <th>...</th>\n",
       "      <th>statezip_WA 98198</th>\n",
       "      <th>statezip_WA 98199</th>\n",
       "      <th>statezip_WA 98288</th>\n",
       "      <th>statezip_WA 98354</th>\n",
       "      <th>country_USA</th>\n",
       "      <th>condition_1</th>\n",
       "      <th>condition_2</th>\n",
       "      <th>condition_3</th>\n",
       "      <th>condition_4</th>\n",
       "      <th>condition_5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.833514</td>\n",
       "      <td>3.130000e+05</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.50</td>\n",
       "      <td>1340</td>\n",
       "      <td>7912</td>\n",
       "      <td>1.5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1340</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.833514</td>\n",
       "      <td>2.384000e+06</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.50</td>\n",
       "      <td>3650</td>\n",
       "      <td>9050</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>3370</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-1.833514</td>\n",
       "      <td>3.420000e+05</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.00</td>\n",
       "      <td>1930</td>\n",
       "      <td>11947</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1930</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-1.833514</td>\n",
       "      <td>4.200000e+05</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.25</td>\n",
       "      <td>2000</td>\n",
       "      <td>8030</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1000</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.833514</td>\n",
       "      <td>5.500000e+05</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.50</td>\n",
       "      <td>1940</td>\n",
       "      <td>10500</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1140</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4595</th>\n",
       "      <td>1.616831</td>\n",
       "      <td>3.081667e+05</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.75</td>\n",
       "      <td>1510</td>\n",
       "      <td>6360</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1510</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4596</th>\n",
       "      <td>1.616831</td>\n",
       "      <td>5.343333e+05</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.50</td>\n",
       "      <td>1460</td>\n",
       "      <td>7573</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1460</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4597</th>\n",
       "      <td>1.616831</td>\n",
       "      <td>4.169042e+05</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.50</td>\n",
       "      <td>3010</td>\n",
       "      <td>7014</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3010</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4598</th>\n",
       "      <td>1.667571</td>\n",
       "      <td>2.034000e+05</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.00</td>\n",
       "      <td>2090</td>\n",
       "      <td>6630</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1070</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4599</th>\n",
       "      <td>1.667571</td>\n",
       "      <td>2.206000e+05</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.50</td>\n",
       "      <td>1490</td>\n",
       "      <td>8102</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1490</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4600 rows × 4665 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          date         price  bedrooms  bathrooms  sqft_living  sqft_lot  \\\n",
       "0    -1.833514  3.130000e+05       3.0       1.50         1340      7912   \n",
       "1    -1.833514  2.384000e+06       5.0       2.50         3650      9050   \n",
       "2    -1.833514  3.420000e+05       3.0       2.00         1930     11947   \n",
       "3    -1.833514  4.200000e+05       3.0       2.25         2000      8030   \n",
       "4    -1.833514  5.500000e+05       4.0       2.50         1940     10500   \n",
       "...        ...           ...       ...        ...          ...       ...   \n",
       "4595  1.616831  3.081667e+05       3.0       1.75         1510      6360   \n",
       "4596  1.616831  5.343333e+05       3.0       2.50         1460      7573   \n",
       "4597  1.616831  4.169042e+05       3.0       2.50         3010      7014   \n",
       "4598  1.667571  2.034000e+05       4.0       2.00         2090      6630   \n",
       "4599  1.667571  2.206000e+05       3.0       2.50         1490      8102   \n",
       "\n",
       "      floors  waterfront  view  sqft_above  ...  statezip_WA 98198  \\\n",
       "0        1.5           0     0        1340  ...                  0   \n",
       "1        2.0           0     4        3370  ...                  0   \n",
       "2        1.0           0     0        1930  ...                  0   \n",
       "3        1.0           0     0        1000  ...                  0   \n",
       "4        1.0           0     0        1140  ...                  0   \n",
       "...      ...         ...   ...         ...  ...                ...   \n",
       "4595     1.0           0     0        1510  ...                  0   \n",
       "4596     2.0           0     0        1460  ...                  0   \n",
       "4597     2.0           0     0        3010  ...                  0   \n",
       "4598     1.0           0     0        1070  ...                  0   \n",
       "4599     2.0           0     0        1490  ...                  0   \n",
       "\n",
       "      statezip_WA 98199  statezip_WA 98288  statezip_WA 98354  country_USA  \\\n",
       "0                     0                  0                  0            1   \n",
       "1                     0                  0                  0            1   \n",
       "2                     0                  0                  0            1   \n",
       "3                     0                  0                  0            1   \n",
       "4                     0                  0                  0            1   \n",
       "...                 ...                ...                ...          ...   \n",
       "4595                  0                  0                  0            1   \n",
       "4596                  0                  0                  0            1   \n",
       "4597                  0                  0                  0            1   \n",
       "4598                  0                  0                  0            1   \n",
       "4599                  0                  0                  0            1   \n",
       "\n",
       "      condition_1  condition_2  condition_3  condition_4  condition_5  \n",
       "0               0            0            1            0            0  \n",
       "1               0            0            0            0            1  \n",
       "2               0            0            0            1            0  \n",
       "3               0            0            0            1            0  \n",
       "4               0            0            0            1            0  \n",
       "...           ...          ...          ...          ...          ...  \n",
       "4595            0            0            0            1            0  \n",
       "4596            0            0            1            0            0  \n",
       "4597            0            0            1            0            0  \n",
       "4598            0            0            1            0            0  \n",
       "4599            0            0            0            1            0  \n",
       "\n",
       "[4600 rows x 4665 columns]"
      ]
     },
     "execution_count": 138,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Objects to Numbers\n",
    "df = raw_df.copy()\n",
    "CATEGORICAL_COLS = [\"street\", \"city\", \"statezip\", \"country\", \"condition\"]\n",
    "for col in CATEGORICAL_COLS:\n",
    "    one_hot_df = pd.get_dummies(df[col], prefix=col)\n",
    "    df = pd.concat([\n",
    "        df.drop(columns=col),\n",
    "        one_hot_df\n",
    "    ], axis=1)\n",
    "df[\"date\"] = pd.to_datetime(df[\"date\"]).apply(lambda x: x.value)\n",
    "df[\"date\"] = (df[\"date\"] - df[\"date\"].mean()) / df[\"date\"].std()  # Normalize \n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "date             0\n",
      "price            0\n",
      "bedrooms         0\n",
      "bathrooms        0\n",
      "sqft_living      0\n",
      "sqft_lot         0\n",
      "floors           0\n",
      "waterfront       0\n",
      "view             0\n",
      "condition        0\n",
      "sqft_above       0\n",
      "sqft_basement    0\n",
      "yr_built         0\n",
      "yr_renovated     0\n",
      "street           0\n",
      "city             0\n",
      "statezip         0\n",
      "country          0\n",
      "dtype: int64\n",
      "date                0\n",
      "price              49\n",
      "bedrooms            2\n",
      "bathrooms           2\n",
      "sqft_living         0\n",
      "sqft_lot            0\n",
      "floors              0\n",
      "waterfront       4567\n",
      "view             4140\n",
      "condition           0\n",
      "sqft_above          0\n",
      "sqft_basement    2745\n",
      "yr_built            0\n",
      "yr_renovated     2735\n",
      "street              0\n",
      "city                0\n",
      "statezip            0\n",
      "country             0\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# Missing Values\n",
    "print(raw_df.isnull().sum())\n",
    "print(raw_df[raw_df==0].count())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Mohsen\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:2: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  \n",
      "C:\\Users\\Mohsen\\Anaconda3\\lib\\site-packages\\pandas\\core\\generic.py:6392: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  return self._update_inplace(result)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>date</th>\n",
       "      <th>price</th>\n",
       "      <th>bedrooms</th>\n",
       "      <th>bathrooms</th>\n",
       "      <th>sqft_living</th>\n",
       "      <th>sqft_lot</th>\n",
       "      <th>floors</th>\n",
       "      <th>waterfront</th>\n",
       "      <th>view</th>\n",
       "      <th>sqft_above</th>\n",
       "      <th>...</th>\n",
       "      <th>statezip_WA 98198</th>\n",
       "      <th>statezip_WA 98199</th>\n",
       "      <th>statezip_WA 98288</th>\n",
       "      <th>statezip_WA 98354</th>\n",
       "      <th>country_USA</th>\n",
       "      <th>condition_1</th>\n",
       "      <th>condition_2</th>\n",
       "      <th>condition_3</th>\n",
       "      <th>condition_4</th>\n",
       "      <th>condition_5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.833514</td>\n",
       "      <td>3.130000e+05</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.50</td>\n",
       "      <td>1340</td>\n",
       "      <td>7912</td>\n",
       "      <td>1.5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1340</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.833514</td>\n",
       "      <td>2.384000e+06</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.50</td>\n",
       "      <td>3650</td>\n",
       "      <td>9050</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>3370</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-1.833514</td>\n",
       "      <td>3.420000e+05</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.00</td>\n",
       "      <td>1930</td>\n",
       "      <td>11947</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1930</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-1.833514</td>\n",
       "      <td>4.200000e+05</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.25</td>\n",
       "      <td>2000</td>\n",
       "      <td>8030</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1000</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.833514</td>\n",
       "      <td>5.500000e+05</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.50</td>\n",
       "      <td>1940</td>\n",
       "      <td>10500</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1140</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4595</th>\n",
       "      <td>1.616831</td>\n",
       "      <td>3.081667e+05</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.75</td>\n",
       "      <td>1510</td>\n",
       "      <td>6360</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1510</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4596</th>\n",
       "      <td>1.616831</td>\n",
       "      <td>5.343333e+05</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.50</td>\n",
       "      <td>1460</td>\n",
       "      <td>7573</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1460</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4597</th>\n",
       "      <td>1.616831</td>\n",
       "      <td>4.169042e+05</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.50</td>\n",
       "      <td>3010</td>\n",
       "      <td>7014</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3010</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4598</th>\n",
       "      <td>1.667571</td>\n",
       "      <td>2.034000e+05</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.00</td>\n",
       "      <td>2090</td>\n",
       "      <td>6630</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1070</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4599</th>\n",
       "      <td>1.667571</td>\n",
       "      <td>2.206000e+05</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.50</td>\n",
       "      <td>1490</td>\n",
       "      <td>8102</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1490</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4551 rows × 4665 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          date         price  bedrooms  bathrooms  sqft_living  sqft_lot  \\\n",
       "0    -1.833514  3.130000e+05       3.0       1.50         1340      7912   \n",
       "1    -1.833514  2.384000e+06       5.0       2.50         3650      9050   \n",
       "2    -1.833514  3.420000e+05       3.0       2.00         1930     11947   \n",
       "3    -1.833514  4.200000e+05       3.0       2.25         2000      8030   \n",
       "4    -1.833514  5.500000e+05       4.0       2.50         1940     10500   \n",
       "...        ...           ...       ...        ...          ...       ...   \n",
       "4595  1.616831  3.081667e+05       3.0       1.75         1510      6360   \n",
       "4596  1.616831  5.343333e+05       3.0       2.50         1460      7573   \n",
       "4597  1.616831  4.169042e+05       3.0       2.50         3010      7014   \n",
       "4598  1.667571  2.034000e+05       4.0       2.00         2090      6630   \n",
       "4599  1.667571  2.206000e+05       3.0       2.50         1490      8102   \n",
       "\n",
       "      floors  waterfront  view  sqft_above  ...  statezip_WA 98198  \\\n",
       "0        1.5           0     0        1340  ...                  0   \n",
       "1        2.0           0     4        3370  ...                  0   \n",
       "2        1.0           0     0        1930  ...                  0   \n",
       "3        1.0           0     0        1000  ...                  0   \n",
       "4        1.0           0     0        1140  ...                  0   \n",
       "...      ...         ...   ...         ...  ...                ...   \n",
       "4595     1.0           0     0        1510  ...                  0   \n",
       "4596     2.0           0     0        1460  ...                  0   \n",
       "4597     2.0           0     0        3010  ...                  0   \n",
       "4598     1.0           0     0        1070  ...                  0   \n",
       "4599     2.0           0     0        1490  ...                  0   \n",
       "\n",
       "      statezip_WA 98199  statezip_WA 98288  statezip_WA 98354  country_USA  \\\n",
       "0                     0                  0                  0            1   \n",
       "1                     0                  0                  0            1   \n",
       "2                     0                  0                  0            1   \n",
       "3                     0                  0                  0            1   \n",
       "4                     0                  0                  0            1   \n",
       "...                 ...                ...                ...          ...   \n",
       "4595                  0                  0                  0            1   \n",
       "4596                  0                  0                  0            1   \n",
       "4597                  0                  0                  0            1   \n",
       "4598                  0                  0                  0            1   \n",
       "4599                  0                  0                  0            1   \n",
       "\n",
       "      condition_1  condition_2  condition_3  condition_4  condition_5  \n",
       "0               0            0            1            0            0  \n",
       "1               0            0            0            0            1  \n",
       "2               0            0            0            1            0  \n",
       "3               0            0            0            1            0  \n",
       "4               0            0            0            1            0  \n",
       "...           ...          ...          ...          ...          ...  \n",
       "4595            0            0            0            1            0  \n",
       "4596            0            0            1            0            0  \n",
       "4597            0            0            1            0            0  \n",
       "4598            0            0            1            0            0  \n",
       "4599            0            0            0            1            0  \n",
       "\n",
       "[4551 rows x 4665 columns]"
      ]
     },
     "execution_count": 140,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = df.loc[df['price'] != 0]\n",
    "df['yr_renovated'] = df['yr_renovated'].replace(0, np.NaN)\n",
    "df['yr_renovated'].fillna((df['yr_renovated'].mean()), inplace=True)\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "date                         0\n",
      "price                        0\n",
      "bedrooms                     2\n",
      "bathrooms                    2\n",
      "sqft_living                  0\n",
      "sqft_lot                     0\n",
      "floors                       0\n",
      "waterfront                4521\n",
      "view                      4103\n",
      "sqft_above                   0\n",
      "sqft_basement             2718\n",
      "yr_built                     0\n",
      "yr_renovated                 0\n",
      "street_1 View Ln NE       4550\n",
      "street_10 W Etruria St    4550\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "print(df[df==0].count()[:15])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((4551, 4664), (4551,))"
      ]
     },
     "execution_count": 142,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x = df.drop(columns='price').to_numpy()\n",
    "y = df['price'].to_numpy()\n",
    "x.shape, y.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train, x_test, y_train, y_test = train_test_split(x, np.log(y), test_size=0.2, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "metadata": {},
   "outputs": [],
   "source": [
    "class TrainerRegression:\n",
    "    def __init__(self, units=1024, activation_function=\"relu\", \n",
    "                 optimizer=\"adam\", n_layers=4, loss_function='mean_squared_error') -> None:\n",
    "        self.optimizer = optimizer\n",
    "        self.units = units\n",
    "        self.activation_function = activation_function\n",
    "        self.n_layers = n_layers\n",
    "        self.loss_function = loss_function\n",
    "        self.model = self.build_mlp_model()\n",
    "        self.history = None\n",
    "        self.training_time = None\n",
    "    \n",
    "    def print_summary(self):\n",
    "        print(\"######################\")\n",
    "        print(f\"Training Time: {self.training_time:.2f}s\")\n",
    "        print(\"batch size:\", self.batch_size)\n",
    "        print(\"n_layers:\", self.n_layers)\n",
    "        print(\"optimizer:\", self.optimizer)\n",
    "        print(\"activation_function:\", self.activation_function)\n",
    "        print(\"loss_function\", self.loss_function)\n",
    "        print(self.model.summary())\n",
    "\n",
    "    def build_mlp_model(self):\n",
    "        model = tf.keras.Sequential([\n",
    "#             tf.keras.layers.Input(shape=(4664, )),\n",
    "            tf.keras.layers.Normalization()\n",
    "        ])\n",
    "        model.add(tf.keras.layers.LayerNormalization())\n",
    "        for n in range(self.n_layers):\n",
    "            model.add(tf.keras.layers.Dense(self.units / 2**n))\n",
    "            model.add(tf.keras.layers.LayerNormalization())\n",
    "            model.add(tf.keras.layers.Activation(self.activation_function))\n",
    "            model.add(tf.keras.layers.Dropout(0.1))\n",
    "        model.add(tf.keras.layers.Dense(1))\n",
    "        \n",
    "        model.compile(\n",
    "            optimizer=self.optimizer,\n",
    "            loss=self.loss_function,\n",
    "            metrics=['mean_squared_error', 'mean_absolute_error']\n",
    "        )\n",
    "        return model\n",
    "    \n",
    "    def train(self, x_train, y_train, batch_size=32, epochs=50):\n",
    "        self.batch_size = batch_size\n",
    "        early_stopping = tf.keras.callbacks.EarlyStopping(\n",
    "            monitor='val_loss', \n",
    "            verbose=1,\n",
    "            patience=10,\n",
    "            mode='min',\n",
    "            restore_best_weights=True\n",
    "        )\n",
    "        start = time()\n",
    "        self.history = self.model.fit(\n",
    "            x_train, y_train,\n",
    "            batch_size=batch_size,\n",
    "            epochs=epochs,\n",
    "            verbose=1,\n",
    "            validation_split=0.15,\n",
    "            callbacks=[early_stopping],\n",
    "        )\n",
    "        self.training_time = time() - start\n",
    "    \n",
    "    def plot_history(self):\n",
    "        fig = plt.figure(figsize=(14, 3))\n",
    "        metrics = ['loss', 'mean_squared_error', 'mean_absolute_error']\n",
    "        for n, metric in enumerate(metrics):\n",
    "            plt.subplot(1, 3, n+1)\n",
    "            plt.plot(self.history.epoch, self.history.history[metric], label='Train')\n",
    "            plt.plot(self.history.epoch, self.history.history[f\"val_{metric}\"], linestyle=\"--\", label='Validation')\n",
    "            plt.xlabel('Epoch')\n",
    "            plt.ylabel(metric)\n",
    "            plt.title(metric)\n",
    "        plt.legend()\n",
    "        plt.show()\n",
    "\n",
    "    \n",
    "    def evaluate(self, x_test, y_test):\n",
    "        [test_loss, test_mse, test_mae] = self.model.evaluate(x_test, y_test)\n",
    "        print(\"Test Loss:\", test_loss, \n",
    "              \"Test MSE:\", test_mse,\n",
    "              \"Test MAE:\", test_mae\n",
    "        )\n",
    "        preds = self.model.predict(x_test)\n",
    "        plt.title('Targets vs. Predictions')\n",
    "        plt.scatter(y_test, preds)\n",
    "        plt.xlabel('Target')\n",
    "        plt.ylabel('Prediction')\n",
    "        plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/200\n",
      "49/49 [==============================] - 1s 15ms/step - loss: 7.4895 - mean_squared_error: 7.4895 - mean_absolute_error: 1.6381 - val_loss: 0.3258 - val_mean_squared_error: 0.3258 - val_mean_absolute_error: 0.4497\n",
      "Epoch 2/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.3262 - mean_squared_error: 0.3262 - mean_absolute_error: 0.4435 - val_loss: 0.2006 - val_mean_squared_error: 0.2006 - val_mean_absolute_error: 0.3460\n",
      "Epoch 3/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2714 - mean_squared_error: 0.2714 - mean_absolute_error: 0.4073 - val_loss: 0.1864 - val_mean_squared_error: 0.1864 - val_mean_absolute_error: 0.3332\n",
      "Epoch 4/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.2594 - mean_squared_error: 0.2594 - mean_absolute_error: 0.3936 - val_loss: 0.1908 - val_mean_squared_error: 0.1908 - val_mean_absolute_error: 0.3535\n",
      "Epoch 5/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2389 - mean_squared_error: 0.2389 - mean_absolute_error: 0.3765 - val_loss: 0.1580 - val_mean_squared_error: 0.1580 - val_mean_absolute_error: 0.3108\n",
      "Epoch 6/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.2494 - mean_squared_error: 0.2494 - mean_absolute_error: 0.3865 - val_loss: 0.1670 - val_mean_squared_error: 0.1670 - val_mean_absolute_error: 0.3300\n",
      "Epoch 7/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2292 - mean_squared_error: 0.2292 - mean_absolute_error: 0.3697 - val_loss: 0.1518 - val_mean_squared_error: 0.1518 - val_mean_absolute_error: 0.3051\n",
      "Epoch 8/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2296 - mean_squared_error: 0.2296 - mean_absolute_error: 0.3701 - val_loss: 0.1504 - val_mean_squared_error: 0.1504 - val_mean_absolute_error: 0.3044\n",
      "Epoch 9/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2194 - mean_squared_error: 0.2194 - mean_absolute_error: 0.3631 - val_loss: 0.1475 - val_mean_squared_error: 0.1475 - val_mean_absolute_error: 0.3038\n",
      "Epoch 10/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.2187 - mean_squared_error: 0.2187 - mean_absolute_error: 0.3645 - val_loss: 0.1495 - val_mean_squared_error: 0.1495 - val_mean_absolute_error: 0.3074\n",
      "Epoch 11/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.2230 - mean_squared_error: 0.2230 - mean_absolute_error: 0.3673 - val_loss: 0.1527 - val_mean_squared_error: 0.1527 - val_mean_absolute_error: 0.3017\n",
      "Epoch 12/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2199 - mean_squared_error: 0.2199 - mean_absolute_error: 0.3621 - val_loss: 0.1469 - val_mean_squared_error: 0.1469 - val_mean_absolute_error: 0.2994\n",
      "Epoch 13/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.2167 - mean_squared_error: 0.2167 - mean_absolute_error: 0.3574 - val_loss: 0.1492 - val_mean_squared_error: 0.1492 - val_mean_absolute_error: 0.3048\n",
      "Epoch 14/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.2115 - mean_squared_error: 0.2115 - mean_absolute_error: 0.3585 - val_loss: 0.1494 - val_mean_squared_error: 0.1494 - val_mean_absolute_error: 0.3003\n",
      "Epoch 15/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.2163 - mean_squared_error: 0.2163 - mean_absolute_error: 0.3585 - val_loss: 0.1475 - val_mean_squared_error: 0.1475 - val_mean_absolute_error: 0.3027\n",
      "Epoch 16/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.2129 - mean_squared_error: 0.2129 - mean_absolute_error: 0.3560 - val_loss: 0.1536 - val_mean_squared_error: 0.1536 - val_mean_absolute_error: 0.3174\n",
      "Epoch 17/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2065 - mean_squared_error: 0.2065 - mean_absolute_error: 0.3517 - val_loss: 0.1464 - val_mean_squared_error: 0.1464 - val_mean_absolute_error: 0.3065\n",
      "Epoch 18/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.2119 - mean_squared_error: 0.2119 - mean_absolute_error: 0.3570 - val_loss: 0.1695 - val_mean_squared_error: 0.1695 - val_mean_absolute_error: 0.3163\n",
      "Epoch 19/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2240 - mean_squared_error: 0.2240 - mean_absolute_error: 0.3677 - val_loss: 0.1438 - val_mean_squared_error: 0.1438 - val_mean_absolute_error: 0.2958\n",
      "Epoch 20/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.2055 - mean_squared_error: 0.2055 - mean_absolute_error: 0.3496 - val_loss: 0.1745 - val_mean_squared_error: 0.1745 - val_mean_absolute_error: 0.3183\n",
      "Epoch 21/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.2131 - mean_squared_error: 0.2131 - mean_absolute_error: 0.3560 - val_loss: 0.1714 - val_mean_squared_error: 0.1714 - val_mean_absolute_error: 0.3174\n",
      "Epoch 22/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.1996 - mean_squared_error: 0.1996 - mean_absolute_error: 0.3471 - val_loss: 0.1749 - val_mean_squared_error: 0.1749 - val_mean_absolute_error: 0.3414\n",
      "Epoch 23/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.2046 - mean_squared_error: 0.2046 - mean_absolute_error: 0.3493 - val_loss: 0.1470 - val_mean_squared_error: 0.1470 - val_mean_absolute_error: 0.2989\n",
      "Epoch 24/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.1933 - mean_squared_error: 0.1933 - mean_absolute_error: 0.3384 - val_loss: 0.1429 - val_mean_squared_error: 0.1429 - val_mean_absolute_error: 0.2958\n",
      "Epoch 25/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.1933 - mean_squared_error: 0.1933 - mean_absolute_error: 0.3384 - val_loss: 0.1433 - val_mean_squared_error: 0.1433 - val_mean_absolute_error: 0.2956\n",
      "Epoch 26/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.1982 - mean_squared_error: 0.1982 - mean_absolute_error: 0.3464 - val_loss: 0.1592 - val_mean_squared_error: 0.1592 - val_mean_absolute_error: 0.3242\n",
      "Epoch 27/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.1981 - mean_squared_error: 0.1981 - mean_absolute_error: 0.3427 - val_loss: 0.1437 - val_mean_squared_error: 0.1437 - val_mean_absolute_error: 0.3034\n",
      "Epoch 28/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.1921 - mean_squared_error: 0.1921 - mean_absolute_error: 0.3396 - val_loss: 0.1788 - val_mean_squared_error: 0.1788 - val_mean_absolute_error: 0.3468\n",
      "Epoch 29/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2102 - mean_squared_error: 0.2102 - mean_absolute_error: 0.3534 - val_loss: 0.1409 - val_mean_squared_error: 0.1409 - val_mean_absolute_error: 0.2981\n",
      "Epoch 30/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.2000 - mean_squared_error: 0.2000 - mean_absolute_error: 0.3444 - val_loss: 0.1556 - val_mean_squared_error: 0.1556 - val_mean_absolute_error: 0.3186\n",
      "Epoch 31/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.2007 - mean_squared_error: 0.2007 - mean_absolute_error: 0.3450 - val_loss: 0.1462 - val_mean_squared_error: 0.1462 - val_mean_absolute_error: 0.3070\n",
      "Epoch 32/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.1884 - mean_squared_error: 0.1884 - mean_absolute_error: 0.3341 - val_loss: 0.1456 - val_mean_squared_error: 0.1456 - val_mean_absolute_error: 0.3059\n",
      "Epoch 33/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.1889 - mean_squared_error: 0.1889 - mean_absolute_error: 0.3332 - val_loss: 0.1489 - val_mean_squared_error: 0.1489 - val_mean_absolute_error: 0.3104\n",
      "Epoch 34/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.1847 - mean_squared_error: 0.1847 - mean_absolute_error: 0.3317 - val_loss: 0.1437 - val_mean_squared_error: 0.1437 - val_mean_absolute_error: 0.2956\n",
      "Epoch 35/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.2157 - mean_squared_error: 0.2157 - mean_absolute_error: 0.3608 - val_loss: 0.1404 - val_mean_squared_error: 0.1404 - val_mean_absolute_error: 0.2934\n",
      "Epoch 36/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.1945 - mean_squared_error: 0.1945 - mean_absolute_error: 0.3410 - val_loss: 0.1698 - val_mean_squared_error: 0.1698 - val_mean_absolute_error: 0.3141\n",
      "Epoch 37/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.1981 - mean_squared_error: 0.1981 - mean_absolute_error: 0.3422 - val_loss: 0.1410 - val_mean_squared_error: 0.1410 - val_mean_absolute_error: 0.2979\n",
      "Epoch 38/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.1923 - mean_squared_error: 0.1923 - mean_absolute_error: 0.3350 - val_loss: 0.1428 - val_mean_squared_error: 0.1428 - val_mean_absolute_error: 0.2936\n",
      "Epoch 39/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.1780 - mean_squared_error: 0.1780 - mean_absolute_error: 0.3243 - val_loss: 0.2393 - val_mean_squared_error: 0.2393 - val_mean_absolute_error: 0.3860\n",
      "Epoch 40/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2041 - mean_squared_error: 0.2041 - mean_absolute_error: 0.3484 - val_loss: 0.1400 - val_mean_squared_error: 0.1400 - val_mean_absolute_error: 0.2964\n",
      "Epoch 41/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.1819 - mean_squared_error: 0.1819 - mean_absolute_error: 0.3284 - val_loss: 0.1713 - val_mean_squared_error: 0.1713 - val_mean_absolute_error: 0.3376\n",
      "Epoch 42/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.1917 - mean_squared_error: 0.1917 - mean_absolute_error: 0.3333 - val_loss: 0.2082 - val_mean_squared_error: 0.2082 - val_mean_absolute_error: 0.3514\n",
      "Epoch 43/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.1978 - mean_squared_error: 0.1978 - mean_absolute_error: 0.3446 - val_loss: 0.1508 - val_mean_squared_error: 0.1508 - val_mean_absolute_error: 0.3128\n",
      "Epoch 44/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.1829 - mean_squared_error: 0.1829 - mean_absolute_error: 0.3296 - val_loss: 0.2024 - val_mean_squared_error: 0.2024 - val_mean_absolute_error: 0.3444\n",
      "Epoch 45/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.2073 - mean_squared_error: 0.2073 - mean_absolute_error: 0.3510 - val_loss: 0.1576 - val_mean_squared_error: 0.1576 - val_mean_absolute_error: 0.3227\n",
      "Epoch 46/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.1808 - mean_squared_error: 0.1808 - mean_absolute_error: 0.3292 - val_loss: 0.1745 - val_mean_squared_error: 0.1745 - val_mean_absolute_error: 0.3192\n",
      "Epoch 47/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.1950 - mean_squared_error: 0.1950 - mean_absolute_error: 0.3440 - val_loss: 0.1458 - val_mean_squared_error: 0.1458 - val_mean_absolute_error: 0.3075\n",
      "Epoch 48/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.1902 - mean_squared_error: 0.1902 - mean_absolute_error: 0.3385 - val_loss: 0.1422 - val_mean_squared_error: 0.1422 - val_mean_absolute_error: 0.2987\n",
      "Epoch 49/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 0.1824 - mean_squared_error: 0.1824 - mean_absolute_error: 0.3294 - val_loss: 0.1500 - val_mean_squared_error: 0.1500 - val_mean_absolute_error: 0.3119\n",
      "Epoch 50/200\n",
      "43/49 [=========================>....] - ETA: 0s - loss: 0.1852 - mean_squared_error: 0.1852 - mean_absolute_error: 0.3297Restoring model weights from the end of the best epoch: 40.\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.1805 - mean_squared_error: 0.1805 - mean_absolute_error: 0.3266 - val_loss: 0.1522 - val_mean_squared_error: 0.1522 - val_mean_absolute_error: 0.3000\n",
      "Epoch 50: early stopping\n",
      "######################\n",
      "Training Time: 21.50s\n",
      "batch size: 64\n",
      "n_layers: 1\n",
      "optimizer: <keras.optimizer_v2.adam.Adam object at 0x000001CBA65032B0>\n",
      "activation_function: relu\n",
      "loss_function mean_squared_error\n",
      "Model: \"sequential_114\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " normalization_113 (Normaliz  (None, 4664)             9329      \n",
      " ation)                                                          \n",
      "                                                                 \n",
      " layer_normalization_427 (La  (None, 4664)             9328      \n",
      " yerNormalization)                                               \n",
      "                                                                 \n",
      " dense_566 (Dense)           (None, 2048)              9553920   \n",
      "                                                                 \n",
      " layer_normalization_428 (La  (None, 2048)             4096      \n",
      " yerNormalization)                                               \n",
      "                                                                 \n",
      " activation_353 (Activation)  (None, 2048)             0         \n",
      "                                                                 \n",
      " dropout_452 (Dropout)       (None, 2048)              0         \n",
      "                                                                 \n",
      " dense_567 (Dense)           (None, 1)                 2049      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 9,578,722\n",
      "Trainable params: 9,569,393\n",
      "Non-trainable params: 9,329\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "29/29 [==============================] - 0s 3ms/step - loss: 0.1410 - mean_squared_error: 0.1410 - mean_absolute_error: 0.2970\n",
      "Test Loss: 0.14103904366493225 Test MSE: 0.14103904366493225 Test MAE: 0.2970487177371979\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for n_layer in [3]:\n",
    "    trainer_reg = TrainerRegression(\n",
    "        units=2048,\n",
    "        activation_function=\"relu\", \n",
    "        optimizer=tf.keras.optimizers.Adam(learning_rate=1e-3), \n",
    "        n_layers=n_layer\n",
    "    )\n",
    "    trainer_reg.train(x_train, y_train, \n",
    "                      batch_size=64,\n",
    "                      epochs=200)\n",
    "    trainer_reg.print_summary()\n",
    "    trainer_reg.plot_history()\n",
    "    trainer_reg.evaluate(x_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/200\n",
      "49/49 [==============================] - 2s 18ms/step - loss: 4.3272 - mean_squared_error: 4.3272 - mean_absolute_error: 0.7879 - val_loss: 0.2011 - val_mean_squared_error: 0.2011 - val_mean_absolute_error: 0.3612\n",
      "Epoch 2/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2957 - mean_squared_error: 0.2957 - mean_absolute_error: 0.4203 - val_loss: 0.2197 - val_mean_squared_error: 0.2197 - val_mean_absolute_error: 0.3580\n",
      "Epoch 3/200\n",
      "49/49 [==============================] - 1s 11ms/step - loss: 0.2612 - mean_squared_error: 0.2612 - mean_absolute_error: 0.3985 - val_loss: 0.1554 - val_mean_squared_error: 0.1554 - val_mean_absolute_error: 0.3123\n",
      "Epoch 4/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2418 - mean_squared_error: 0.2418 - mean_absolute_error: 0.3831 - val_loss: 0.1519 - val_mean_squared_error: 0.1519 - val_mean_absolute_error: 0.3031\n",
      "Epoch 5/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2360 - mean_squared_error: 0.2360 - mean_absolute_error: 0.3800 - val_loss: 0.1555 - val_mean_squared_error: 0.1555 - val_mean_absolute_error: 0.3036\n",
      "Epoch 6/200\n",
      "49/49 [==============================] - 1s 10ms/step - loss: 0.2289 - mean_squared_error: 0.2289 - mean_absolute_error: 0.3692 - val_loss: 0.1483 - val_mean_squared_error: 0.1483 - val_mean_absolute_error: 0.3001\n",
      "Epoch 7/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2184 - mean_squared_error: 0.2184 - mean_absolute_error: 0.3636 - val_loss: 0.1530 - val_mean_squared_error: 0.1530 - val_mean_absolute_error: 0.3114\n",
      "Epoch 8/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2300 - mean_squared_error: 0.2300 - mean_absolute_error: 0.3729 - val_loss: 0.1740 - val_mean_squared_error: 0.1740 - val_mean_absolute_error: 0.3170\n",
      "Epoch 9/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2301 - mean_squared_error: 0.2301 - mean_absolute_error: 0.3734 - val_loss: 0.1932 - val_mean_squared_error: 0.1932 - val_mean_absolute_error: 0.3341\n",
      "Epoch 10/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2236 - mean_squared_error: 0.2236 - mean_absolute_error: 0.3670 - val_loss: 0.1804 - val_mean_squared_error: 0.1804 - val_mean_absolute_error: 0.3477\n",
      "Epoch 11/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2160 - mean_squared_error: 0.2160 - mean_absolute_error: 0.3586 - val_loss: 0.1970 - val_mean_squared_error: 0.1970 - val_mean_absolute_error: 0.3645\n",
      "Epoch 12/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2277 - mean_squared_error: 0.2277 - mean_absolute_error: 0.3703 - val_loss: 0.1455 - val_mean_squared_error: 0.1455 - val_mean_absolute_error: 0.2983\n",
      "Epoch 13/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2092 - mean_squared_error: 0.2092 - mean_absolute_error: 0.3539 - val_loss: 0.1467 - val_mean_squared_error: 0.1467 - val_mean_absolute_error: 0.3047\n",
      "Epoch 14/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2233 - mean_squared_error: 0.2233 - mean_absolute_error: 0.3644 - val_loss: 0.1528 - val_mean_squared_error: 0.1528 - val_mean_absolute_error: 0.3025\n",
      "Epoch 15/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2158 - mean_squared_error: 0.2158 - mean_absolute_error: 0.3574 - val_loss: 0.1507 - val_mean_squared_error: 0.1507 - val_mean_absolute_error: 0.3082\n",
      "Epoch 16/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2049 - mean_squared_error: 0.2049 - mean_absolute_error: 0.3517 - val_loss: 0.1466 - val_mean_squared_error: 0.1466 - val_mean_absolute_error: 0.3061\n",
      "Epoch 17/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2108 - mean_squared_error: 0.2108 - mean_absolute_error: 0.3570 - val_loss: 0.1480 - val_mean_squared_error: 0.1480 - val_mean_absolute_error: 0.2986\n",
      "Epoch 18/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.1976 - mean_squared_error: 0.1976 - mean_absolute_error: 0.3461 - val_loss: 0.1668 - val_mean_squared_error: 0.1668 - val_mean_absolute_error: 0.3122\n",
      "Epoch 19/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2086 - mean_squared_error: 0.2086 - mean_absolute_error: 0.3517 - val_loss: 0.1469 - val_mean_squared_error: 0.1469 - val_mean_absolute_error: 0.2982\n",
      "Epoch 20/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2120 - mean_squared_error: 0.2120 - mean_absolute_error: 0.3577 - val_loss: 0.1478 - val_mean_squared_error: 0.1478 - val_mean_absolute_error: 0.3045\n",
      "Epoch 21/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2061 - mean_squared_error: 0.2061 - mean_absolute_error: 0.3511 - val_loss: 0.1783 - val_mean_squared_error: 0.1783 - val_mean_absolute_error: 0.3205\n",
      "Epoch 22/200\n",
      "49/49 [==============================] - ETA: 0s - loss: 0.2056 - mean_squared_error: 0.2056 - mean_absolute_error: 0.3485Restoring model weights from the end of the best epoch: 12.\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2056 - mean_squared_error: 0.2056 - mean_absolute_error: 0.3485 - val_loss: 0.1609 - val_mean_squared_error: 0.1609 - val_mean_absolute_error: 0.3235\n",
      "Epoch 22: early stopping\n",
      "######################\n",
      "Training Time: 11.66s\n",
      "batch size: 64\n",
      "n_layers: 2\n",
      "optimizer: <keras.optimizer_v2.adam.Adam object at 0x000001CBA6004390>\n",
      "activation_function: relu\n",
      "loss_function mean_squared_error\n",
      "Model: \"sequential_125\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " normalization_124 (Normaliz  (None, 4664)             9329      \n",
      " ation)                                                          \n",
      "                                                                 \n",
      " layer_normalization_469 (La  (None, 4664)             9328      \n",
      " yerNormalization)                                               \n",
      "                                                                 \n",
      " dense_608 (Dense)           (None, 2048)              9553920   \n",
      "                                                                 \n",
      " layer_normalization_470 (La  (None, 2048)             4096      \n",
      " yerNormalization)                                               \n",
      "                                                                 \n",
      " activation_384 (Activation)  (None, 2048)             0         \n",
      "                                                                 \n",
      " dropout_483 (Dropout)       (None, 2048)              0         \n",
      "                                                                 \n",
      " dense_609 (Dense)           (None, 1024)              2098176   \n",
      "                                                                 \n",
      " layer_normalization_471 (La  (None, 1024)             2048      \n",
      " yerNormalization)                                               \n",
      "                                                                 \n",
      " activation_385 (Activation)  (None, 1024)             0         \n",
      "                                                                 \n",
      " dropout_484 (Dropout)       (None, 1024)              0         \n",
      "                                                                 \n",
      " dense_610 (Dense)           (None, 1)                 1025      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 11,677,922\n",
      "Trainable params: 11,668,593\n",
      "Non-trainable params: 9,329\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "29/29 [==============================] - 0s 3ms/step - loss: 0.1445 - mean_squared_error: 0.1445 - mean_absolute_error: 0.2999\n",
      "Test Loss: 0.14445987343788147 Test MSE: 0.14445987343788147 Test MAE: 0.29986485838890076\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/200\n",
      "49/49 [==============================] - 2s 15ms/step - loss: 5.0361 - mean_squared_error: 5.0361 - mean_absolute_error: 1.1167 - val_loss: 0.3425 - val_mean_squared_error: 0.3425 - val_mean_absolute_error: 0.4633\n",
      "Epoch 2/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.3793 - mean_squared_error: 0.3793 - mean_absolute_error: 0.4841 - val_loss: 0.3012 - val_mean_squared_error: 0.3012 - val_mean_absolute_error: 0.4330\n",
      "Epoch 3/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.3947 - mean_squared_error: 0.3947 - mean_absolute_error: 0.4859 - val_loss: 0.3029 - val_mean_squared_error: 0.3029 - val_mean_absolute_error: 0.4337\n",
      "Epoch 4/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.3988 - mean_squared_error: 0.3988 - mean_absolute_error: 0.4896 - val_loss: 0.3022 - val_mean_squared_error: 0.3022 - val_mean_absolute_error: 0.4332\n",
      "Epoch 5/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.3797 - mean_squared_error: 0.3797 - mean_absolute_error: 0.4819 - val_loss: 0.3002 - val_mean_squared_error: 0.3002 - val_mean_absolute_error: 0.4324\n",
      "Epoch 6/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.3702 - mean_squared_error: 0.3702 - mean_absolute_error: 0.4743 - val_loss: 0.3053 - val_mean_squared_error: 0.3053 - val_mean_absolute_error: 0.4366\n",
      "Epoch 7/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.3744 - mean_squared_error: 0.3744 - mean_absolute_error: 0.4781 - val_loss: 0.2983 - val_mean_squared_error: 0.2983 - val_mean_absolute_error: 0.4309\n",
      "Epoch 8/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.3957 - mean_squared_error: 0.3957 - mean_absolute_error: 0.4893 - val_loss: 0.2991 - val_mean_squared_error: 0.2991 - val_mean_absolute_error: 0.4309\n",
      "Epoch 9/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.3783 - mean_squared_error: 0.3783 - mean_absolute_error: 0.4819 - val_loss: 0.2972 - val_mean_squared_error: 0.2972 - val_mean_absolute_error: 0.4303\n",
      "Epoch 10/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.3672 - mean_squared_error: 0.3672 - mean_absolute_error: 0.4689 - val_loss: 0.2905 - val_mean_squared_error: 0.2905 - val_mean_absolute_error: 0.4247\n",
      "Epoch 11/200\n",
      "49/49 [==============================] - 1s 11ms/step - loss: 0.3626 - mean_squared_error: 0.3626 - mean_absolute_error: 0.4710 - val_loss: 0.2869 - val_mean_squared_error: 0.2869 - val_mean_absolute_error: 0.4218\n",
      "Epoch 12/200\n",
      "49/49 [==============================] - 1s 11ms/step - loss: 0.3637 - mean_squared_error: 0.3637 - mean_absolute_error: 0.4684 - val_loss: 0.2731 - val_mean_squared_error: 0.2731 - val_mean_absolute_error: 0.4115\n",
      "Epoch 13/200\n",
      "49/49 [==============================] - 1s 10ms/step - loss: 0.3324 - mean_squared_error: 0.3324 - mean_absolute_error: 0.4472 - val_loss: 0.2557 - val_mean_squared_error: 0.2557 - val_mean_absolute_error: 0.3946\n",
      "Epoch 14/200\n",
      "49/49 [==============================] - 1s 11ms/step - loss: 0.3041 - mean_squared_error: 0.3041 - mean_absolute_error: 0.4277 - val_loss: 0.2190 - val_mean_squared_error: 0.2190 - val_mean_absolute_error: 0.3612\n",
      "Epoch 15/200\n",
      "49/49 [==============================] - 1s 11ms/step - loss: 0.2828 - mean_squared_error: 0.2828 - mean_absolute_error: 0.4139 - val_loss: 0.1774 - val_mean_squared_error: 0.1774 - val_mean_absolute_error: 0.3301\n",
      "Epoch 16/200\n",
      "49/49 [==============================] - 1s 11ms/step - loss: 0.2552 - mean_squared_error: 0.2552 - mean_absolute_error: 0.3887 - val_loss: 0.1714 - val_mean_squared_error: 0.1714 - val_mean_absolute_error: 0.3173\n",
      "Epoch 17/200\n",
      "49/49 [==============================] - 1s 10ms/step - loss: 0.2617 - mean_squared_error: 0.2617 - mean_absolute_error: 0.3969 - val_loss: 0.1651 - val_mean_squared_error: 0.1651 - val_mean_absolute_error: 0.3247\n",
      "Epoch 18/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2449 - mean_squared_error: 0.2449 - mean_absolute_error: 0.3847 - val_loss: 0.1544 - val_mean_squared_error: 0.1544 - val_mean_absolute_error: 0.3052\n",
      "Epoch 19/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2477 - mean_squared_error: 0.2477 - mean_absolute_error: 0.3846 - val_loss: 0.1587 - val_mean_squared_error: 0.1587 - val_mean_absolute_error: 0.3185\n",
      "Epoch 20/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2405 - mean_squared_error: 0.2405 - mean_absolute_error: 0.3811 - val_loss: 0.1540 - val_mean_squared_error: 0.1540 - val_mean_absolute_error: 0.3126\n",
      "Epoch 21/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2332 - mean_squared_error: 0.2332 - mean_absolute_error: 0.3758 - val_loss: 0.1518 - val_mean_squared_error: 0.1518 - val_mean_absolute_error: 0.3093\n",
      "Epoch 22/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2316 - mean_squared_error: 0.2316 - mean_absolute_error: 0.3733 - val_loss: 0.1698 - val_mean_squared_error: 0.1698 - val_mean_absolute_error: 0.3138\n",
      "Epoch 23/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2290 - mean_squared_error: 0.2290 - mean_absolute_error: 0.3731 - val_loss: 0.1507 - val_mean_squared_error: 0.1507 - val_mean_absolute_error: 0.3082\n",
      "Epoch 24/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2278 - mean_squared_error: 0.2278 - mean_absolute_error: 0.3763 - val_loss: 0.1524 - val_mean_squared_error: 0.1524 - val_mean_absolute_error: 0.3050\n",
      "Epoch 25/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2356 - mean_squared_error: 0.2356 - mean_absolute_error: 0.3770 - val_loss: 0.1505 - val_mean_squared_error: 0.1505 - val_mean_absolute_error: 0.3073\n",
      "Epoch 26/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2329 - mean_squared_error: 0.2329 - mean_absolute_error: 0.3724 - val_loss: 0.1551 - val_mean_squared_error: 0.1551 - val_mean_absolute_error: 0.3141\n",
      "Epoch 27/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2226 - mean_squared_error: 0.2226 - mean_absolute_error: 0.3626 - val_loss: 0.1509 - val_mean_squared_error: 0.1509 - val_mean_absolute_error: 0.3079\n",
      "Epoch 28/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2370 - mean_squared_error: 0.2370 - mean_absolute_error: 0.3825 - val_loss: 0.1604 - val_mean_squared_error: 0.1604 - val_mean_absolute_error: 0.3065\n",
      "Epoch 29/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2315 - mean_squared_error: 0.2315 - mean_absolute_error: 0.3766 - val_loss: 0.1512 - val_mean_squared_error: 0.1512 - val_mean_absolute_error: 0.3034\n",
      "Epoch 30/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2191 - mean_squared_error: 0.2191 - mean_absolute_error: 0.3626 - val_loss: 0.1564 - val_mean_squared_error: 0.1564 - val_mean_absolute_error: 0.3185\n",
      "Epoch 31/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2267 - mean_squared_error: 0.2267 - mean_absolute_error: 0.3688 - val_loss: 0.1553 - val_mean_squared_error: 0.1553 - val_mean_absolute_error: 0.3174\n",
      "Epoch 32/200\n",
      "49/49 [==============================] - 1s 10ms/step - loss: 0.2316 - mean_squared_error: 0.2316 - mean_absolute_error: 0.3748 - val_loss: 0.1490 - val_mean_squared_error: 0.1490 - val_mean_absolute_error: 0.3057\n",
      "Epoch 33/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2157 - mean_squared_error: 0.2157 - mean_absolute_error: 0.3597 - val_loss: 0.1530 - val_mean_squared_error: 0.1530 - val_mean_absolute_error: 0.3025\n",
      "Epoch 34/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2358 - mean_squared_error: 0.2358 - mean_absolute_error: 0.3762 - val_loss: 0.1528 - val_mean_squared_error: 0.1528 - val_mean_absolute_error: 0.3054\n",
      "Epoch 35/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2163 - mean_squared_error: 0.2163 - mean_absolute_error: 0.3566 - val_loss: 0.1536 - val_mean_squared_error: 0.1536 - val_mean_absolute_error: 0.3151\n",
      "Epoch 36/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2165 - mean_squared_error: 0.2165 - mean_absolute_error: 0.3606 - val_loss: 0.1532 - val_mean_squared_error: 0.1532 - val_mean_absolute_error: 0.3028\n",
      "Epoch 37/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2274 - mean_squared_error: 0.2274 - mean_absolute_error: 0.3685 - val_loss: 0.1487 - val_mean_squared_error: 0.1487 - val_mean_absolute_error: 0.3050\n",
      "Epoch 38/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2180 - mean_squared_error: 0.2180 - mean_absolute_error: 0.3639 - val_loss: 0.1534 - val_mean_squared_error: 0.1534 - val_mean_absolute_error: 0.3122\n",
      "Epoch 39/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2219 - mean_squared_error: 0.2219 - mean_absolute_error: 0.3641 - val_loss: 0.1547 - val_mean_squared_error: 0.1547 - val_mean_absolute_error: 0.3035\n",
      "Epoch 40/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2221 - mean_squared_error: 0.2221 - mean_absolute_error: 0.3654 - val_loss: 0.1574 - val_mean_squared_error: 0.1574 - val_mean_absolute_error: 0.3063\n",
      "Epoch 41/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2342 - mean_squared_error: 0.2342 - mean_absolute_error: 0.3773 - val_loss: 0.1732 - val_mean_squared_error: 0.1732 - val_mean_absolute_error: 0.3168\n",
      "Epoch 42/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2161 - mean_squared_error: 0.2161 - mean_absolute_error: 0.3600 - val_loss: 0.1494 - val_mean_squared_error: 0.1494 - val_mean_absolute_error: 0.3076\n",
      "Epoch 43/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2148 - mean_squared_error: 0.2148 - mean_absolute_error: 0.3580 - val_loss: 0.1588 - val_mean_squared_error: 0.1588 - val_mean_absolute_error: 0.3222\n",
      "Epoch 44/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1996 - mean_squared_error: 0.1996 - mean_absolute_error: 0.3448 - val_loss: 0.1511 - val_mean_squared_error: 0.1511 - val_mean_absolute_error: 0.3020\n",
      "Epoch 45/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2105 - mean_squared_error: 0.2105 - mean_absolute_error: 0.3538 - val_loss: 0.1577 - val_mean_squared_error: 0.1577 - val_mean_absolute_error: 0.3046\n",
      "Epoch 46/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2090 - mean_squared_error: 0.2090 - mean_absolute_error: 0.3524 - val_loss: 0.1603 - val_mean_squared_error: 0.1603 - val_mean_absolute_error: 0.3085\n",
      "Epoch 47/200\n",
      "49/49 [==============================] - 1s 10ms/step - loss: 0.2139 - mean_squared_error: 0.2139 - mean_absolute_error: 0.3567 - val_loss: 0.1473 - val_mean_squared_error: 0.1473 - val_mean_absolute_error: 0.3045\n",
      "Epoch 48/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2109 - mean_squared_error: 0.2109 - mean_absolute_error: 0.3528 - val_loss: 0.1524 - val_mean_squared_error: 0.1524 - val_mean_absolute_error: 0.3141\n",
      "Epoch 49/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2036 - mean_squared_error: 0.2036 - mean_absolute_error: 0.3483 - val_loss: 0.1534 - val_mean_squared_error: 0.1534 - val_mean_absolute_error: 0.3027\n",
      "Epoch 50/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2151 - mean_squared_error: 0.2151 - mean_absolute_error: 0.3565 - val_loss: 0.1568 - val_mean_squared_error: 0.1568 - val_mean_absolute_error: 0.3199\n",
      "Epoch 51/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2050 - mean_squared_error: 0.2050 - mean_absolute_error: 0.3497 - val_loss: 0.1520 - val_mean_squared_error: 0.1520 - val_mean_absolute_error: 0.3040\n",
      "Epoch 52/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2078 - mean_squared_error: 0.2078 - mean_absolute_error: 0.3501 - val_loss: 0.1491 - val_mean_squared_error: 0.1491 - val_mean_absolute_error: 0.3093\n",
      "Epoch 53/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2040 - mean_squared_error: 0.2040 - mean_absolute_error: 0.3484 - val_loss: 0.1487 - val_mean_squared_error: 0.1487 - val_mean_absolute_error: 0.3080\n",
      "Epoch 54/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2122 - mean_squared_error: 0.2122 - mean_absolute_error: 0.3551 - val_loss: 0.1475 - val_mean_squared_error: 0.1475 - val_mean_absolute_error: 0.3057\n",
      "Epoch 55/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2065 - mean_squared_error: 0.2065 - mean_absolute_error: 0.3509 - val_loss: 0.1516 - val_mean_squared_error: 0.1516 - val_mean_absolute_error: 0.3128\n",
      "Epoch 56/200\n",
      "49/49 [==============================] - 1s 11ms/step - loss: 0.2022 - mean_squared_error: 0.2022 - mean_absolute_error: 0.3488 - val_loss: 0.1471 - val_mean_squared_error: 0.1471 - val_mean_absolute_error: 0.3062\n",
      "Epoch 57/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1971 - mean_squared_error: 0.1971 - mean_absolute_error: 0.3427 - val_loss: 0.1464 - val_mean_squared_error: 0.1464 - val_mean_absolute_error: 0.3005\n",
      "Epoch 58/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2011 - mean_squared_error: 0.2011 - mean_absolute_error: 0.3454 - val_loss: 0.1546 - val_mean_squared_error: 0.1546 - val_mean_absolute_error: 0.3176\n",
      "Epoch 59/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2069 - mean_squared_error: 0.2069 - mean_absolute_error: 0.3543 - val_loss: 0.1471 - val_mean_squared_error: 0.1471 - val_mean_absolute_error: 0.3028\n",
      "Epoch 60/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2003 - mean_squared_error: 0.2003 - mean_absolute_error: 0.3447 - val_loss: 0.1696 - val_mean_squared_error: 0.1696 - val_mean_absolute_error: 0.3142\n",
      "Epoch 61/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2013 - mean_squared_error: 0.2013 - mean_absolute_error: 0.3486 - val_loss: 0.1492 - val_mean_squared_error: 0.1492 - val_mean_absolute_error: 0.3099\n",
      "Epoch 62/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2062 - mean_squared_error: 0.2062 - mean_absolute_error: 0.3541 - val_loss: 0.1558 - val_mean_squared_error: 0.1558 - val_mean_absolute_error: 0.3026\n",
      "Epoch 63/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1996 - mean_squared_error: 0.1996 - mean_absolute_error: 0.3436 - val_loss: 0.1540 - val_mean_squared_error: 0.1540 - val_mean_absolute_error: 0.3024\n",
      "Epoch 64/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2004 - mean_squared_error: 0.2004 - mean_absolute_error: 0.3435 - val_loss: 0.1464 - val_mean_squared_error: 0.1464 - val_mean_absolute_error: 0.3019\n",
      "Epoch 65/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.1961 - mean_squared_error: 0.1961 - mean_absolute_error: 0.3425 - val_loss: 0.1475 - val_mean_squared_error: 0.1475 - val_mean_absolute_error: 0.3085\n",
      "Epoch 66/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1991 - mean_squared_error: 0.1991 - mean_absolute_error: 0.3426 - val_loss: 0.1459 - val_mean_squared_error: 0.1459 - val_mean_absolute_error: 0.3052\n",
      "Epoch 67/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2093 - mean_squared_error: 0.2093 - mean_absolute_error: 0.3547 - val_loss: 0.1532 - val_mean_squared_error: 0.1532 - val_mean_absolute_error: 0.3159\n",
      "Epoch 68/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2082 - mean_squared_error: 0.2082 - mean_absolute_error: 0.3506 - val_loss: 0.1567 - val_mean_squared_error: 0.1567 - val_mean_absolute_error: 0.3045\n",
      "Epoch 69/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1970 - mean_squared_error: 0.1970 - mean_absolute_error: 0.3403 - val_loss: 0.1498 - val_mean_squared_error: 0.1498 - val_mean_absolute_error: 0.3018\n",
      "Epoch 70/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1967 - mean_squared_error: 0.1967 - mean_absolute_error: 0.3421 - val_loss: 0.1636 - val_mean_squared_error: 0.1636 - val_mean_absolute_error: 0.3094\n",
      "Epoch 71/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1926 - mean_squared_error: 0.1926 - mean_absolute_error: 0.3388 - val_loss: 0.1470 - val_mean_squared_error: 0.1470 - val_mean_absolute_error: 0.3005\n",
      "Epoch 72/200\n",
      "49/49 [==============================] - 1s 11ms/step - loss: 0.1924 - mean_squared_error: 0.1924 - mean_absolute_error: 0.3360 - val_loss: 0.1443 - val_mean_squared_error: 0.1443 - val_mean_absolute_error: 0.3030\n",
      "Epoch 73/200\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2060 - mean_squared_error: 0.2060 - mean_absolute_error: 0.3501 - val_loss: 0.1464 - val_mean_squared_error: 0.1464 - val_mean_absolute_error: 0.3065\n",
      "Epoch 74/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1976 - mean_squared_error: 0.1976 - mean_absolute_error: 0.3431 - val_loss: 0.1621 - val_mean_squared_error: 0.1621 - val_mean_absolute_error: 0.3078\n",
      "Epoch 75/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1955 - mean_squared_error: 0.1955 - mean_absolute_error: 0.3435 - val_loss: 0.1464 - val_mean_squared_error: 0.1464 - val_mean_absolute_error: 0.3069\n",
      "Epoch 76/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1968 - mean_squared_error: 0.1968 - mean_absolute_error: 0.3442 - val_loss: 0.1446 - val_mean_squared_error: 0.1446 - val_mean_absolute_error: 0.2991\n",
      "Epoch 77/200\n",
      "49/49 [==============================] - 1s 11ms/step - loss: 0.1900 - mean_squared_error: 0.1900 - mean_absolute_error: 0.3352 - val_loss: 0.1433 - val_mean_squared_error: 0.1433 - val_mean_absolute_error: 0.3006\n",
      "Epoch 78/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1846 - mean_squared_error: 0.1846 - mean_absolute_error: 0.3297 - val_loss: 0.1744 - val_mean_squared_error: 0.1744 - val_mean_absolute_error: 0.3416\n",
      "Epoch 79/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1904 - mean_squared_error: 0.1904 - mean_absolute_error: 0.3372 - val_loss: 0.1597 - val_mean_squared_error: 0.1597 - val_mean_absolute_error: 0.3250\n",
      "Epoch 80/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1986 - mean_squared_error: 0.1986 - mean_absolute_error: 0.3393 - val_loss: 0.1434 - val_mean_squared_error: 0.1434 - val_mean_absolute_error: 0.3016\n",
      "Epoch 81/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1930 - mean_squared_error: 0.1930 - mean_absolute_error: 0.3369 - val_loss: 0.1441 - val_mean_squared_error: 0.1441 - val_mean_absolute_error: 0.2994\n",
      "Epoch 82/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1968 - mean_squared_error: 0.1968 - mean_absolute_error: 0.3440 - val_loss: 0.1790 - val_mean_squared_error: 0.1790 - val_mean_absolute_error: 0.3224\n",
      "Epoch 83/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1964 - mean_squared_error: 0.1964 - mean_absolute_error: 0.3385 - val_loss: 0.1725 - val_mean_squared_error: 0.1725 - val_mean_absolute_error: 0.3164\n",
      "Epoch 84/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2070 - mean_squared_error: 0.2070 - mean_absolute_error: 0.3514 - val_loss: 0.1708 - val_mean_squared_error: 0.1708 - val_mean_absolute_error: 0.3143\n",
      "Epoch 85/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1906 - mean_squared_error: 0.1906 - mean_absolute_error: 0.3342 - val_loss: 0.1668 - val_mean_squared_error: 0.1668 - val_mean_absolute_error: 0.3105\n",
      "Epoch 86/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.1925 - mean_squared_error: 0.1925 - mean_absolute_error: 0.3422 - val_loss: 0.1529 - val_mean_squared_error: 0.1529 - val_mean_absolute_error: 0.3156\n",
      "Epoch 87/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1814 - mean_squared_error: 0.1814 - mean_absolute_error: 0.3275 - val_loss: 0.1430 - val_mean_squared_error: 0.1430 - val_mean_absolute_error: 0.3018\n",
      "Epoch 88/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.1987 - mean_squared_error: 0.1987 - mean_absolute_error: 0.3394 - val_loss: 0.1512 - val_mean_squared_error: 0.1512 - val_mean_absolute_error: 0.2997\n",
      "Epoch 89/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1824 - mean_squared_error: 0.1824 - mean_absolute_error: 0.3282 - val_loss: 0.1427 - val_mean_squared_error: 0.1427 - val_mean_absolute_error: 0.3002\n",
      "Epoch 90/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1802 - mean_squared_error: 0.1802 - mean_absolute_error: 0.3240 - val_loss: 0.1578 - val_mean_squared_error: 0.1578 - val_mean_absolute_error: 0.3225\n",
      "Epoch 91/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1869 - mean_squared_error: 0.1869 - mean_absolute_error: 0.3315 - val_loss: 0.1427 - val_mean_squared_error: 0.1427 - val_mean_absolute_error: 0.3022\n",
      "Epoch 92/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1870 - mean_squared_error: 0.1870 - mean_absolute_error: 0.3320 - val_loss: 0.1691 - val_mean_squared_error: 0.1691 - val_mean_absolute_error: 0.3131\n",
      "Epoch 93/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1887 - mean_squared_error: 0.1887 - mean_absolute_error: 0.3330 - val_loss: 0.1640 - val_mean_squared_error: 0.1640 - val_mean_absolute_error: 0.3298\n",
      "Epoch 94/200\n",
      "49/49 [==============================] - 1s 10ms/step - loss: 0.1926 - mean_squared_error: 0.1926 - mean_absolute_error: 0.3392 - val_loss: 0.1419 - val_mean_squared_error: 0.1419 - val_mean_absolute_error: 0.3007\n",
      "Epoch 95/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.1867 - mean_squared_error: 0.1867 - mean_absolute_error: 0.3347 - val_loss: 0.1435 - val_mean_squared_error: 0.1435 - val_mean_absolute_error: 0.3015\n",
      "Epoch 96/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1862 - mean_squared_error: 0.1862 - mean_absolute_error: 0.3331 - val_loss: 0.1414 - val_mean_squared_error: 0.1414 - val_mean_absolute_error: 0.2987\n",
      "Epoch 97/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1777 - mean_squared_error: 0.1777 - mean_absolute_error: 0.3250 - val_loss: 0.1433 - val_mean_squared_error: 0.1433 - val_mean_absolute_error: 0.2973\n",
      "Epoch 98/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1791 - mean_squared_error: 0.1791 - mean_absolute_error: 0.3247 - val_loss: 0.1442 - val_mean_squared_error: 0.1442 - val_mean_absolute_error: 0.2986\n",
      "Epoch 99/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1780 - mean_squared_error: 0.1780 - mean_absolute_error: 0.3224 - val_loss: 0.1478 - val_mean_squared_error: 0.1478 - val_mean_absolute_error: 0.2978\n",
      "Epoch 100/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.1796 - mean_squared_error: 0.1796 - mean_absolute_error: 0.3270 - val_loss: 0.1480 - val_mean_squared_error: 0.1480 - val_mean_absolute_error: 0.3100\n",
      "Epoch 101/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.1759 - mean_squared_error: 0.1759 - mean_absolute_error: 0.3223 - val_loss: 0.1416 - val_mean_squared_error: 0.1416 - val_mean_absolute_error: 0.2985\n",
      "Epoch 102/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1772 - mean_squared_error: 0.1772 - mean_absolute_error: 0.3225 - val_loss: 0.1572 - val_mean_squared_error: 0.1572 - val_mean_absolute_error: 0.3228\n",
      "Epoch 103/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.1844 - mean_squared_error: 0.1844 - mean_absolute_error: 0.3292 - val_loss: 0.2412 - val_mean_squared_error: 0.2412 - val_mean_absolute_error: 0.3842\n",
      "Epoch 104/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1922 - mean_squared_error: 0.1922 - mean_absolute_error: 0.3373 - val_loss: 0.1566 - val_mean_squared_error: 0.1566 - val_mean_absolute_error: 0.3036\n",
      "Epoch 105/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.1905 - mean_squared_error: 0.1905 - mean_absolute_error: 0.3333 - val_loss: 0.1742 - val_mean_squared_error: 0.1742 - val_mean_absolute_error: 0.3405\n",
      "Epoch 106/200\n",
      "45/49 [==========================>...] - ETA: 0s - loss: 0.1908 - mean_squared_error: 0.1908 - mean_absolute_error: 0.3386Restoring model weights from the end of the best epoch: 96.\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.1922 - mean_squared_error: 0.1922 - mean_absolute_error: 0.3391 - val_loss: 0.1431 - val_mean_squared_error: 0.1431 - val_mean_absolute_error: 0.2965\n",
      "Epoch 106: early stopping\n",
      "######################\n",
      "Training Time: 52.23s\n",
      "batch size: 64\n",
      "n_layers: 2\n",
      "optimizer: <keras.optimizer_v2.adam.Adam object at 0x000001CBA6096E80>\n",
      "activation_function: sigmoid\n",
      "loss_function mean_squared_error\n",
      "Model: \"sequential_126\"\n",
      "_________________________________________________________________\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " normalization_125 (Normaliz  (None, 4664)             9329      \n",
      " ation)                                                          \n",
      "                                                                 \n",
      " layer_normalization_472 (La  (None, 4664)             9328      \n",
      " yerNormalization)                                               \n",
      "                                                                 \n",
      " dense_611 (Dense)           (None, 2048)              9553920   \n",
      "                                                                 \n",
      " layer_normalization_473 (La  (None, 2048)             4096      \n",
      " yerNormalization)                                               \n",
      "                                                                 \n",
      " activation_386 (Activation)  (None, 2048)             0         \n",
      "                                                                 \n",
      " dropout_485 (Dropout)       (None, 2048)              0         \n",
      "                                                                 \n",
      " dense_612 (Dense)           (None, 1024)              2098176   \n",
      "                                                                 \n",
      " layer_normalization_474 (La  (None, 1024)             2048      \n",
      " yerNormalization)                                               \n",
      "                                                                 \n",
      " activation_387 (Activation)  (None, 1024)             0         \n",
      "                                                                 \n",
      " dropout_486 (Dropout)       (None, 1024)              0         \n",
      "                                                                 \n",
      " dense_613 (Dense)           (None, 1)                 1025      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 11,677,922\n",
      "Trainable params: 11,668,593\n",
      "Non-trainable params: 9,329\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "29/29 [==============================] - 0s 3ms/step - loss: 0.1422 - mean_squared_error: 0.1422 - mean_absolute_error: 0.2968\n",
      "Test Loss: 0.14224666357040405 Test MSE: 0.14224666357040405 Test MAE: 0.2967800199985504\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for activation_function in [\"relu\", \"sigmoid\"]:\n",
    "    trainer_reg = TrainerRegression(\n",
    "        units=2048,\n",
    "        activation_function=activation_function, \n",
    "        optimizer=tf.keras.optimizers.Adam(learning_rate=1e-3), \n",
    "        n_layers=2\n",
    "    )\n",
    "    trainer_reg.train(x_train, y_train, \n",
    "                      batch_size=64,\n",
    "                      epochs=200)\n",
    "    trainer_reg.print_summary()\n",
    "    trainer_reg.plot_history()\n",
    "    trainer_reg.evaluate(x_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/200\n",
      "49/49 [==============================] - 2s 19ms/step - loss: 4.2192 - mean_squared_error: 4.2192 - mean_absolute_error: 0.9859 - val_loss: 0.2372 - val_mean_squared_error: 0.2372 - val_mean_absolute_error: 0.3819\n",
      "Epoch 2/200\n",
      "49/49 [==============================] - 1s 11ms/step - loss: 0.2909 - mean_squared_error: 0.2909 - mean_absolute_error: 0.4188 - val_loss: 0.2170 - val_mean_squared_error: 0.2170 - val_mean_absolute_error: 0.3569\n",
      "Epoch 3/200\n",
      "49/49 [==============================] - 1s 10ms/step - loss: 0.2630 - mean_squared_error: 0.2630 - mean_absolute_error: 0.4010 - val_loss: 0.1953 - val_mean_squared_error: 0.1953 - val_mean_absolute_error: 0.3414\n",
      "Epoch 4/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2378 - mean_squared_error: 0.2378 - mean_absolute_error: 0.3770 - val_loss: 0.1616 - val_mean_squared_error: 0.1616 - val_mean_absolute_error: 0.3209\n",
      "Epoch 5/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2462 - mean_squared_error: 0.2462 - mean_absolute_error: 0.3853 - val_loss: 0.1546 - val_mean_squared_error: 0.1546 - val_mean_absolute_error: 0.3142\n",
      "Epoch 6/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2370 - mean_squared_error: 0.2370 - mean_absolute_error: 0.3753 - val_loss: 0.1493 - val_mean_squared_error: 0.1493 - val_mean_absolute_error: 0.3027\n",
      "Epoch 7/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2327 - mean_squared_error: 0.2327 - mean_absolute_error: 0.3735 - val_loss: 0.1809 - val_mean_squared_error: 0.1809 - val_mean_absolute_error: 0.3462\n",
      "Epoch 8/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2313 - mean_squared_error: 0.2313 - mean_absolute_error: 0.3755 - val_loss: 0.1496 - val_mean_squared_error: 0.1496 - val_mean_absolute_error: 0.3103\n",
      "Epoch 9/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2355 - mean_squared_error: 0.2355 - mean_absolute_error: 0.3757 - val_loss: 0.1493 - val_mean_squared_error: 0.1493 - val_mean_absolute_error: 0.3053\n",
      "Epoch 10/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2198 - mean_squared_error: 0.2198 - mean_absolute_error: 0.3607 - val_loss: 0.1681 - val_mean_squared_error: 0.1681 - val_mean_absolute_error: 0.3121\n",
      "Epoch 11/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2374 - mean_squared_error: 0.2374 - mean_absolute_error: 0.3799 - val_loss: 0.1631 - val_mean_squared_error: 0.1631 - val_mean_absolute_error: 0.3293\n",
      "Epoch 12/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2438 - mean_squared_error: 0.2438 - mean_absolute_error: 0.3794 - val_loss: 0.1548 - val_mean_squared_error: 0.1548 - val_mean_absolute_error: 0.3188\n",
      "Epoch 13/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2211 - mean_squared_error: 0.2211 - mean_absolute_error: 0.3678 - val_loss: 0.1493 - val_mean_squared_error: 0.1493 - val_mean_absolute_error: 0.3011\n",
      "Epoch 14/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2109 - mean_squared_error: 0.2109 - mean_absolute_error: 0.3555 - val_loss: 0.1653 - val_mean_squared_error: 0.1653 - val_mean_absolute_error: 0.3110\n",
      "Epoch 15/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2114 - mean_squared_error: 0.2114 - mean_absolute_error: 0.3544 - val_loss: 0.1437 - val_mean_squared_error: 0.1437 - val_mean_absolute_error: 0.2999\n",
      "Epoch 16/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2106 - mean_squared_error: 0.2106 - mean_absolute_error: 0.3562 - val_loss: 0.1894 - val_mean_squared_error: 0.1894 - val_mean_absolute_error: 0.3303\n",
      "Epoch 17/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2077 - mean_squared_error: 0.2077 - mean_absolute_error: 0.3539 - val_loss: 0.1622 - val_mean_squared_error: 0.1622 - val_mean_absolute_error: 0.3057\n",
      "Epoch 18/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2112 - mean_squared_error: 0.2112 - mean_absolute_error: 0.3558 - val_loss: 0.1451 - val_mean_squared_error: 0.1451 - val_mean_absolute_error: 0.2978\n",
      "Epoch 19/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2243 - mean_squared_error: 0.2243 - mean_absolute_error: 0.3671 - val_loss: 0.1467 - val_mean_squared_error: 0.1467 - val_mean_absolute_error: 0.2994\n",
      "Epoch 20/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2113 - mean_squared_error: 0.2113 - mean_absolute_error: 0.3540 - val_loss: 0.1660 - val_mean_squared_error: 0.1660 - val_mean_absolute_error: 0.3321\n",
      "Epoch 21/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2063 - mean_squared_error: 0.2063 - mean_absolute_error: 0.3515 - val_loss: 0.1453 - val_mean_squared_error: 0.1453 - val_mean_absolute_error: 0.3046\n",
      "Epoch 22/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2149 - mean_squared_error: 0.2149 - mean_absolute_error: 0.3598 - val_loss: 0.2343 - val_mean_squared_error: 0.2343 - val_mean_absolute_error: 0.3741\n",
      "Epoch 23/200\n",
      "49/49 [==============================] - 0s 9ms/step - loss: 0.2174 - mean_squared_error: 0.2174 - mean_absolute_error: 0.3630 - val_loss: 0.1725 - val_mean_squared_error: 0.1725 - val_mean_absolute_error: 0.3162\n",
      "Epoch 24/200\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2062 - mean_squared_error: 0.2062 - mean_absolute_error: 0.3529 - val_loss: 0.1469 - val_mean_squared_error: 0.1469 - val_mean_absolute_error: 0.3078\n",
      "Epoch 25/200\n",
      "45/49 [==========================>...] - ETA: 0s - loss: 0.2061 - mean_squared_error: 0.2061 - mean_absolute_error: 0.3481Restoring model weights from the end of the best epoch: 15.\n",
      "49/49 [==============================] - 0s 10ms/step - loss: 0.2072 - mean_squared_error: 0.2072 - mean_absolute_error: 0.3502 - val_loss: 0.1600 - val_mean_squared_error: 0.1600 - val_mean_absolute_error: 0.3046\n",
      "Epoch 25: early stopping\n",
      "######################\n",
      "Training Time: 13.24s\n",
      "batch size: 64\n",
      "n_layers: 2\n",
      "optimizer: <keras.optimizer_v2.adam.Adam object at 0x000001CDBF6D8E80>\n",
      "activation_function: relu\n",
      "loss_function mean_squared_error\n",
      "Model: \"sequential_127\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " normalization_126 (Normaliz  (None, 4664)             9329      \n",
      " ation)                                                          \n",
      "                                                                 \n",
      " layer_normalization_475 (La  (None, 4664)             9328      \n",
      " yerNormalization)                                               \n",
      "                                                                 \n",
      " dense_614 (Dense)           (None, 2048)              9553920   \n",
      "                                                                 \n",
      " layer_normalization_476 (La  (None, 2048)             4096      \n",
      " yerNormalization)                                               \n",
      "                                                                 \n",
      " activation_388 (Activation)  (None, 2048)             0         \n",
      "                                                                 \n",
      " dropout_487 (Dropout)       (None, 2048)              0         \n",
      "                                                                 \n",
      " dense_615 (Dense)           (None, 1024)              2098176   \n",
      "                                                                 \n",
      " layer_normalization_477 (La  (None, 1024)             2048      \n",
      " yerNormalization)                                               \n",
      "                                                                 \n",
      " activation_389 (Activation)  (None, 1024)             0         \n",
      "                                                                 \n",
      " dropout_488 (Dropout)       (None, 1024)              0         \n",
      "                                                                 \n",
      " dense_616 (Dense)           (None, 1)                 1025      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 11,677,922\n",
      "Trainable params: 11,668,593\n",
      "Non-trainable params: 9,329\n",
      "_________________________________________________________________\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "None\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "29/29 [==============================] - 0s 3ms/step - loss: 0.1439 - mean_squared_error: 0.1439 - mean_absolute_error: 0.3015\n",
      "Test Loss: 0.14392122626304626 Test MSE: 0.14392122626304626 Test MAE: 0.30151689052581787\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "trainer_reg = TrainerRegression(\n",
    "    units=2048,\n",
    "    activation_function=\"relu\", \n",
    "    optimizer=tf.keras.optimizers.Adam(learning_rate=1e-3), \n",
    "    n_layers=2,\n",
    "    loss_function='mean_squared_error'\n",
    ")\n",
    "trainer_reg.train(x_train, y_train, \n",
    "                  batch_size=64,\n",
    "                  epochs=200)\n",
    "trainer_reg.print_summary()\n",
    "trainer_reg.plot_history()\n",
    "trainer_reg.evaluate(x_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/200\n",
      "49/49 [==============================] - 1s 11ms/step - loss: 13.0118 - mean_squared_error: 169.6030 - mean_absolute_error: 13.0118 - val_loss: 12.9366 - val_mean_squared_error: 167.6563 - val_mean_absolute_error: 12.9366\n",
      "Epoch 2/200\n",
      "49/49 [==============================] - 0s 8ms/step - loss: 12.8954 - mean_squared_error: 166.5874 - mean_absolute_error: 12.8954 - val_loss: 12.8151 - val_mean_squared_error: 164.5284 - val_mean_absolute_error: 12.8151\n",
      "Epoch 3/200\n",
      "49/49 [==============================] - 0s 7ms/step - loss: 12.7708 - mean_squared_error: 163.3905 - mean_absolute_error: 12.7708 - val_loss: 12.6845 - val_mean_squared_error: 161.1965 - val_mean_absolute_error: 12.6845\n",
      "Epoch 4/200\n",
      "49/49 [==============================] - 0s 7ms/step - loss: 12.6328 - mean_squared_error: 159.8893 - mean_absolute_error: 12.6328 - val_loss: 12.5425 - val_mean_squared_error: 157.6161 - val_mean_absolute_error: 12.5425\n",
      "Epoch 5/200\n",
      "49/49 [==============================] - 0s 7ms/step - loss: 12.4891 - mean_squared_error: 156.2914 - mean_absolute_error: 12.4891 - val_loss: 12.3907 - val_mean_squared_error: 153.8291 - val_mean_absolute_error: 12.3907\n",
      "Epoch 6/200\n",
      "49/49 [==============================] - 0s 7ms/step - loss: 12.3308 - mean_squared_error: 152.3697 - mean_absolute_error: 12.3308 - val_loss: 12.2280 - val_mean_squared_error: 149.8239 - val_mean_absolute_error: 12.2280\n",
      "Epoch 7/200\n",
      "49/49 [==============================] - 0s 7ms/step - loss: 12.1634 - mean_squared_error: 148.2863 - mean_absolute_error: 12.1634 - val_loss: 12.0549 - val_mean_squared_error: 145.6218 - val_mean_absolute_error: 12.0549\n",
      "Epoch 8/200\n",
      "49/49 [==============================] - 0s 7ms/step - loss: 11.9818 - mean_squared_error: 143.9140 - mean_absolute_error: 11.9818 - val_loss: 11.8702 - val_mean_squared_error: 141.2026 - val_mean_absolute_error: 11.8702\n",
      "Epoch 9/200\n",
      "49/49 [==============================] - 0s 7ms/step - loss: 11.7972 - mean_squared_error: 139.5545 - mean_absolute_error: 11.7972 - val_loss: 11.6756 - val_mean_squared_error: 136.6199 - val_mean_absolute_error: 11.6756\n",
      "Epoch 10/200\n",
      "49/49 [==============================] - 0s 7ms/step - loss: 11.5997 - mean_squared_error: 134.9675 - mean_absolute_error: 11.5997 - val_loss: 11.4708 - val_mean_squared_error: 131.8811 - val_mean_absolute_error: 11.4708\n",
      "Epoch 11/200\n",
      "49/49 [==============================] - 0s 7ms/step - loss: 11.3886 - mean_squared_error: 130.1478 - mean_absolute_error: 11.3886 - val_loss: 11.2550 - val_mean_squared_error: 126.9760 - val_mean_absolute_error: 11.2550\n",
      "Epoch 12/200\n",
      "49/49 [==============================] - 0s 7ms/step - loss: 11.1726 - mean_squared_error: 125.3494 - mean_absolute_error: 11.1726 - val_loss: 11.0293 - val_mean_squared_error: 121.9468 - val_mean_absolute_error: 11.0293\n",
      "Epoch 13/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 10.9283 - mean_squared_error: 119.9511 - mean_absolute_error: 10.9283 - val_loss: 10.7917 - val_mean_squared_error: 116.7626 - val_mean_absolute_error: 10.7917\n",
      "Epoch 14/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 10.7042 - mean_squared_error: 115.2353 - mean_absolute_error: 10.7042 - val_loss: 10.5450 - val_mean_squared_error: 111.4989 - val_mean_absolute_error: 10.5450\n",
      "Epoch 15/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 10.4629 - mean_squared_error: 110.1935 - mean_absolute_error: 10.4629 - val_loss: 10.2898 - val_mean_squared_error: 106.1814 - val_mean_absolute_error: 10.2898\n",
      "Epoch 16/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 10.1680 - mean_squared_error: 104.1459 - mean_absolute_error: 10.1680 - val_loss: 10.0213 - val_mean_squared_error: 100.7272 - val_mean_absolute_error: 10.0213\n",
      "Epoch 17/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 9.9015 - mean_squared_error: 98.9203 - mean_absolute_error: 9.9015 - val_loss: 9.7428 - val_mean_squared_error: 95.2220 - val_mean_absolute_error: 9.7428\n",
      "Epoch 18/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 9.6186 - mean_squared_error: 93.5291 - mean_absolute_error: 9.6186 - val_loss: 9.4544 - val_mean_squared_error: 89.6857 - val_mean_absolute_error: 9.4544\n",
      "Epoch 19/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 9.3443 - mean_squared_error: 88.5716 - mean_absolute_error: 9.3443 - val_loss: 9.1579 - val_mean_squared_error: 84.1686 - val_mean_absolute_error: 9.1579\n",
      "Epoch 20/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 9.0252 - mean_squared_error: 82.8025 - mean_absolute_error: 9.0252 - val_loss: 8.8498 - val_mean_squared_error: 78.6193 - val_mean_absolute_error: 8.8498\n",
      "Epoch 21/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 8.7252 - mean_squared_error: 77.7117 - mean_absolute_error: 8.7252 - val_loss: 8.5328 - val_mean_squared_error: 73.1089 - val_mean_absolute_error: 8.5328\n",
      "Epoch 22/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 8.4129 - mean_squared_error: 72.6020 - mean_absolute_error: 8.4129 - val_loss: 8.2075 - val_mean_squared_error: 67.6647 - val_mean_absolute_error: 8.2075\n",
      "Epoch 23/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 8.0457 - mean_squared_error: 66.6740 - mean_absolute_error: 8.0457 - val_loss: 7.8711 - val_mean_squared_error: 62.2548 - val_mean_absolute_error: 7.8711\n",
      "Epoch 24/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 7.7057 - mean_squared_error: 61.5878 - mean_absolute_error: 7.7057 - val_loss: 7.5244 - val_mean_squared_error: 56.9169 - val_mean_absolute_error: 7.5244\n",
      "Epoch 25/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 7.3719 - mean_squared_error: 56.9290 - mean_absolute_error: 7.3719 - val_loss: 7.1688 - val_mean_squared_error: 51.6925 - val_mean_absolute_error: 7.1688\n",
      "Epoch 26/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 6.9899 - mean_squared_error: 51.6450 - mean_absolute_error: 6.9899 - val_loss: 6.8047 - val_mean_squared_error: 46.6049 - val_mean_absolute_error: 6.8047\n",
      "Epoch 27/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 6.6482 - mean_squared_error: 47.4307 - mean_absolute_error: 6.6482 - val_loss: 6.4321 - val_mean_squared_error: 41.6730 - val_mean_absolute_error: 6.4321\n",
      "Epoch 28/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 6.2382 - mean_squared_error: 42.3899 - mean_absolute_error: 6.2382 - val_loss: 6.0499 - val_mean_squared_error: 36.9020 - val_mean_absolute_error: 6.0499\n",
      "Epoch 29/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 5.8797 - mean_squared_error: 38.5834 - mean_absolute_error: 5.8797 - val_loss: 5.6603 - val_mean_squared_error: 32.3395 - val_mean_absolute_error: 5.6603\n",
      "Epoch 30/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 5.4589 - mean_squared_error: 34.2234 - mean_absolute_error: 5.4589 - val_loss: 5.2609 - val_mean_squared_error: 27.9784 - val_mean_absolute_error: 5.2609\n",
      "Epoch 31/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 5.0488 - mean_squared_error: 30.2931 - mean_absolute_error: 5.0488 - val_loss: 4.8525 - val_mean_squared_error: 23.8478 - val_mean_absolute_error: 4.8525\n",
      "Epoch 32/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 4.6497 - mean_squared_error: 27.0146 - mean_absolute_error: 4.6497 - val_loss: 4.4359 - val_mean_squared_error: 19.9783 - val_mean_absolute_error: 4.4359\n",
      "Epoch 33/200\n",
      "49/49 [==============================] - 0s 7ms/step - loss: 4.2719 - mean_squared_error: 24.5756 - mean_absolute_error: 4.2719 - val_loss: 4.0137 - val_mean_squared_error: 16.4104 - val_mean_absolute_error: 4.0137\n",
      "Epoch 34/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 3.7751 - mean_squared_error: 20.6094 - mean_absolute_error: 3.7751 - val_loss: 3.5817 - val_mean_squared_error: 13.1291 - val_mean_absolute_error: 3.5817\n",
      "Epoch 35/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 3.3684 - mean_squared_error: 18.6597 - mean_absolute_error: 3.3684 - val_loss: 3.1407 - val_mean_squared_error: 10.1650 - val_mean_absolute_error: 3.1407\n",
      "Epoch 36/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 2.9508 - mean_squared_error: 16.8582 - mean_absolute_error: 2.9508 - val_loss: 2.6940 - val_mean_squared_error: 7.5586 - val_mean_absolute_error: 2.6940\n",
      "Epoch 37/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 2.5045 - mean_squared_error: 15.2410 - mean_absolute_error: 2.5045 - val_loss: 2.2408 - val_mean_squared_error: 5.3219 - val_mean_absolute_error: 2.2408\n",
      "Epoch 38/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.9849 - mean_squared_error: 13.1099 - mean_absolute_error: 1.9849 - val_loss: 1.7924 - val_mean_squared_error: 3.5135 - val_mean_absolute_error: 1.7924\n",
      "Epoch 39/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.6840 - mean_squared_error: 13.0532 - mean_absolute_error: 1.6840 - val_loss: 1.4301 - val_mean_squared_error: 2.3380 - val_mean_absolute_error: 1.4301\n",
      "Epoch 40/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.6159 - mean_squared_error: 13.8798 - mean_absolute_error: 1.6159 - val_loss: 1.2390 - val_mean_squared_error: 1.8190 - val_mean_absolute_error: 1.2390\n",
      "Epoch 41/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4897 - mean_squared_error: 12.6760 - mean_absolute_error: 1.4897 - val_loss: 1.1542 - val_mean_squared_error: 1.6106 - val_mean_absolute_error: 1.1542\n",
      "Epoch 42/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.5774 - mean_squared_error: 13.6935 - mean_absolute_error: 1.5774 - val_loss: 1.1324 - val_mean_squared_error: 1.5595 - val_mean_absolute_error: 1.1324\n",
      "Epoch 43/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.5219 - mean_squared_error: 13.0681 - mean_absolute_error: 1.5219 - val_loss: 1.1214 - val_mean_squared_error: 1.5339 - val_mean_absolute_error: 1.1214\n",
      "Epoch 44/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4474 - mean_squared_error: 12.2205 - mean_absolute_error: 1.4474 - val_loss: 1.1157 - val_mean_squared_error: 1.5208 - val_mean_absolute_error: 1.1157\n",
      "Epoch 45/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.5404 - mean_squared_error: 13.3394 - mean_absolute_error: 1.5404 - val_loss: 1.1191 - val_mean_squared_error: 1.5287 - val_mean_absolute_error: 1.1191\n",
      "Epoch 46/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.5301 - mean_squared_error: 13.0788 - mean_absolute_error: 1.5301 - val_loss: 1.1185 - val_mean_squared_error: 1.5274 - val_mean_absolute_error: 1.1185\n",
      "Epoch 47/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4402 - mean_squared_error: 12.0562 - mean_absolute_error: 1.4402 - val_loss: 1.1163 - val_mean_squared_error: 1.5222 - val_mean_absolute_error: 1.1163\n",
      "Epoch 48/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4627 - mean_squared_error: 12.3042 - mean_absolute_error: 1.4627 - val_loss: 1.1167 - val_mean_squared_error: 1.5232 - val_mean_absolute_error: 1.1167\n",
      "Epoch 49/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4746 - mean_squared_error: 12.4238 - mean_absolute_error: 1.4746 - val_loss: 1.1150 - val_mean_squared_error: 1.5192 - val_mean_absolute_error: 1.1150\n",
      "Epoch 50/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4079 - mean_squared_error: 11.7252 - mean_absolute_error: 1.4079 - val_loss: 1.1157 - val_mean_squared_error: 1.5209 - val_mean_absolute_error: 1.1157\n",
      "Epoch 51/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.5383 - mean_squared_error: 13.1906 - mean_absolute_error: 1.5383 - val_loss: 1.1161 - val_mean_squared_error: 1.5216 - val_mean_absolute_error: 1.1161\n",
      "Epoch 52/200\n",
      "49/49 [==============================] - 0s 7ms/step - loss: 1.5606 - mean_squared_error: 13.4295 - mean_absolute_error: 1.5606 - val_loss: 1.1160 - val_mean_squared_error: 1.5215 - val_mean_absolute_error: 1.1160\n",
      "Epoch 53/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.5266 - mean_squared_error: 13.0238 - mean_absolute_error: 1.5266 - val_loss: 1.1062 - val_mean_squared_error: 1.4991 - val_mean_absolute_error: 1.1062\n",
      "Epoch 54/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.5411 - mean_squared_error: 13.2063 - mean_absolute_error: 1.5411 - val_loss: 1.1075 - val_mean_squared_error: 1.5020 - val_mean_absolute_error: 1.1075\n",
      "Epoch 55/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4451 - mean_squared_error: 12.0702 - mean_absolute_error: 1.4451 - val_loss: 1.1079 - val_mean_squared_error: 1.5029 - val_mean_absolute_error: 1.1079\n",
      "Epoch 56/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4092 - mean_squared_error: 11.6743 - mean_absolute_error: 1.4092 - val_loss: 1.1109 - val_mean_squared_error: 1.5097 - val_mean_absolute_error: 1.1109\n",
      "Epoch 57/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4531 - mean_squared_error: 12.1763 - mean_absolute_error: 1.4531 - val_loss: 1.1118 - val_mean_squared_error: 1.5119 - val_mean_absolute_error: 1.1118\n",
      "Epoch 58/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.6417 - mean_squared_error: 14.3148 - mean_absolute_error: 1.6417 - val_loss: 1.1076 - val_mean_squared_error: 1.5023 - val_mean_absolute_error: 1.1076\n",
      "Epoch 59/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3887 - mean_squared_error: 11.3929 - mean_absolute_error: 1.3887 - val_loss: 1.1069 - val_mean_squared_error: 1.5006 - val_mean_absolute_error: 1.1069\n",
      "Epoch 60/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3986 - mean_squared_error: 11.4963 - mean_absolute_error: 1.3986 - val_loss: 1.1058 - val_mean_squared_error: 1.4981 - val_mean_absolute_error: 1.1058\n",
      "Epoch 61/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.5034 - mean_squared_error: 12.6400 - mean_absolute_error: 1.5034 - val_loss: 1.1095 - val_mean_squared_error: 1.5067 - val_mean_absolute_error: 1.1095\n",
      "Epoch 62/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4594 - mean_squared_error: 12.2613 - mean_absolute_error: 1.4594 - val_loss: 1.1110 - val_mean_squared_error: 1.5100 - val_mean_absolute_error: 1.1110\n",
      "Epoch 63/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.5013 - mean_squared_error: 12.6520 - mean_absolute_error: 1.5013 - val_loss: 1.1078 - val_mean_squared_error: 1.5028 - val_mean_absolute_error: 1.1078\n",
      "Epoch 64/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4672 - mean_squared_error: 12.2472 - mean_absolute_error: 1.4672 - val_loss: 1.1064 - val_mean_squared_error: 1.4994 - val_mean_absolute_error: 1.1064\n",
      "Epoch 65/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4627 - mean_squared_error: 12.1877 - mean_absolute_error: 1.4627 - val_loss: 1.1107 - val_mean_squared_error: 1.5092 - val_mean_absolute_error: 1.1107\n",
      "Epoch 66/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3645 - mean_squared_error: 11.1117 - mean_absolute_error: 1.3645 - val_loss: 1.1001 - val_mean_squared_error: 1.4851 - val_mean_absolute_error: 1.1001\n",
      "Epoch 67/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4152 - mean_squared_error: 11.6321 - mean_absolute_error: 1.4152 - val_loss: 1.0996 - val_mean_squared_error: 1.4840 - val_mean_absolute_error: 1.0996\n",
      "Epoch 68/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4924 - mean_squared_error: 12.5599 - mean_absolute_error: 1.4924 - val_loss: 1.0972 - val_mean_squared_error: 1.4785 - val_mean_absolute_error: 1.0972\n",
      "Epoch 69/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3926 - mean_squared_error: 11.3218 - mean_absolute_error: 1.3926 - val_loss: 1.0990 - val_mean_squared_error: 1.4826 - val_mean_absolute_error: 1.0990\n",
      "Epoch 70/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4331 - mean_squared_error: 11.8406 - mean_absolute_error: 1.4331 - val_loss: 1.0971 - val_mean_squared_error: 1.4783 - val_mean_absolute_error: 1.0971\n",
      "Epoch 71/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4702 - mean_squared_error: 12.2565 - mean_absolute_error: 1.4702 - val_loss: 1.1034 - val_mean_squared_error: 1.4927 - val_mean_absolute_error: 1.1034\n",
      "Epoch 72/200\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "49/49 [==============================] - 0s 6ms/step - loss: 1.5089 - mean_squared_error: 12.7136 - mean_absolute_error: 1.5089 - val_loss: 1.1034 - val_mean_squared_error: 1.4925 - val_mean_absolute_error: 1.1034\n",
      "Epoch 73/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.5112 - mean_squared_error: 12.6461 - mean_absolute_error: 1.5112 - val_loss: 1.0936 - val_mean_squared_error: 1.4703 - val_mean_absolute_error: 1.0936\n",
      "Epoch 74/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4143 - mean_squared_error: 11.5775 - mean_absolute_error: 1.4143 - val_loss: 1.0952 - val_mean_squared_error: 1.4741 - val_mean_absolute_error: 1.0952\n",
      "Epoch 75/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3932 - mean_squared_error: 11.2632 - mean_absolute_error: 1.3932 - val_loss: 1.0896 - val_mean_squared_error: 1.4613 - val_mean_absolute_error: 1.0896\n",
      "Epoch 76/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4984 - mean_squared_error: 12.4669 - mean_absolute_error: 1.4984 - val_loss: 1.0873 - val_mean_squared_error: 1.4562 - val_mean_absolute_error: 1.0873\n",
      "Epoch 77/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.5175 - mean_squared_error: 12.7119 - mean_absolute_error: 1.5175 - val_loss: 1.0940 - val_mean_squared_error: 1.4712 - val_mean_absolute_error: 1.0940\n",
      "Epoch 78/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4419 - mean_squared_error: 11.8010 - mean_absolute_error: 1.4419 - val_loss: 1.0937 - val_mean_squared_error: 1.4705 - val_mean_absolute_error: 1.0937\n",
      "Epoch 79/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.5043 - mean_squared_error: 12.5690 - mean_absolute_error: 1.5043 - val_loss: 1.0889 - val_mean_squared_error: 1.4598 - val_mean_absolute_error: 1.0889\n",
      "Epoch 80/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.5180 - mean_squared_error: 12.6816 - mean_absolute_error: 1.5180 - val_loss: 1.0903 - val_mean_squared_error: 1.4630 - val_mean_absolute_error: 1.0903\n",
      "Epoch 81/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4883 - mean_squared_error: 12.3009 - mean_absolute_error: 1.4883 - val_loss: 1.0827 - val_mean_squared_error: 1.4458 - val_mean_absolute_error: 1.0827\n",
      "Epoch 82/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4252 - mean_squared_error: 11.5647 - mean_absolute_error: 1.4252 - val_loss: 1.0827 - val_mean_squared_error: 1.4459 - val_mean_absolute_error: 1.0827\n",
      "Epoch 83/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4972 - mean_squared_error: 12.3707 - mean_absolute_error: 1.4972 - val_loss: 1.0845 - val_mean_squared_error: 1.4499 - val_mean_absolute_error: 1.0845\n",
      "Epoch 84/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4374 - mean_squared_error: 11.7435 - mean_absolute_error: 1.4374 - val_loss: 1.0907 - val_mean_squared_error: 1.4637 - val_mean_absolute_error: 1.0907\n",
      "Epoch 85/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4489 - mean_squared_error: 11.8474 - mean_absolute_error: 1.4489 - val_loss: 1.0866 - val_mean_squared_error: 1.4545 - val_mean_absolute_error: 1.0866\n",
      "Epoch 86/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.5276 - mean_squared_error: 12.7550 - mean_absolute_error: 1.5276 - val_loss: 1.0768 - val_mean_squared_error: 1.4326 - val_mean_absolute_error: 1.0768\n",
      "Epoch 87/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4404 - mean_squared_error: 11.6991 - mean_absolute_error: 1.4404 - val_loss: 1.0794 - val_mean_squared_error: 1.4384 - val_mean_absolute_error: 1.0794\n",
      "Epoch 88/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4499 - mean_squared_error: 11.7686 - mean_absolute_error: 1.4499 - val_loss: 1.0822 - val_mean_squared_error: 1.4446 - val_mean_absolute_error: 1.0822\n",
      "Epoch 89/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4102 - mean_squared_error: 11.3648 - mean_absolute_error: 1.4102 - val_loss: 1.0726 - val_mean_squared_error: 1.4231 - val_mean_absolute_error: 1.0726\n",
      "Epoch 90/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.5199 - mean_squared_error: 12.6326 - mean_absolute_error: 1.5199 - val_loss: 1.0768 - val_mean_squared_error: 1.4325 - val_mean_absolute_error: 1.0768\n",
      "Epoch 91/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4912 - mean_squared_error: 12.3180 - mean_absolute_error: 1.4912 - val_loss: 1.0815 - val_mean_squared_error: 1.4430 - val_mean_absolute_error: 1.0815\n",
      "Epoch 92/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4949 - mean_squared_error: 12.2526 - mean_absolute_error: 1.4949 - val_loss: 1.0766 - val_mean_squared_error: 1.4321 - val_mean_absolute_error: 1.0766\n",
      "Epoch 93/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4744 - mean_squared_error: 11.9439 - mean_absolute_error: 1.4744 - val_loss: 1.0595 - val_mean_squared_error: 1.3942 - val_mean_absolute_error: 1.0595\n",
      "Epoch 94/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4351 - mean_squared_error: 11.5288 - mean_absolute_error: 1.4351 - val_loss: 1.0683 - val_mean_squared_error: 1.4135 - val_mean_absolute_error: 1.0683\n",
      "Epoch 95/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3922 - mean_squared_error: 11.0543 - mean_absolute_error: 1.3922 - val_loss: 1.0665 - val_mean_squared_error: 1.4096 - val_mean_absolute_error: 1.0665\n",
      "Epoch 96/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4690 - mean_squared_error: 11.8613 - mean_absolute_error: 1.4690 - val_loss: 1.0682 - val_mean_squared_error: 1.4134 - val_mean_absolute_error: 1.0682\n",
      "Epoch 97/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3520 - mean_squared_error: 10.5756 - mean_absolute_error: 1.3520 - val_loss: 1.0716 - val_mean_squared_error: 1.4211 - val_mean_absolute_error: 1.0716\n",
      "Epoch 98/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4675 - mean_squared_error: 11.9130 - mean_absolute_error: 1.4675 - val_loss: 1.0695 - val_mean_squared_error: 1.4162 - val_mean_absolute_error: 1.0695\n",
      "Epoch 99/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4085 - mean_squared_error: 11.1694 - mean_absolute_error: 1.4085 - val_loss: 1.0709 - val_mean_squared_error: 1.4194 - val_mean_absolute_error: 1.0709\n",
      "Epoch 100/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3315 - mean_squared_error: 10.2885 - mean_absolute_error: 1.3315 - val_loss: 1.0661 - val_mean_squared_error: 1.4087 - val_mean_absolute_error: 1.0661\n",
      "Epoch 101/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4733 - mean_squared_error: 11.8307 - mean_absolute_error: 1.4733 - val_loss: 1.0569 - val_mean_squared_error: 1.3884 - val_mean_absolute_error: 1.0569\n",
      "Epoch 102/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4333 - mean_squared_error: 11.3732 - mean_absolute_error: 1.4333 - val_loss: 1.0553 - val_mean_squared_error: 1.3849 - val_mean_absolute_error: 1.0553\n",
      "Epoch 103/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4232 - mean_squared_error: 11.2532 - mean_absolute_error: 1.4232 - val_loss: 1.0571 - val_mean_squared_error: 1.3889 - val_mean_absolute_error: 1.0571\n",
      "Epoch 104/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4102 - mean_squared_error: 11.1001 - mean_absolute_error: 1.4102 - val_loss: 1.0581 - val_mean_squared_error: 1.3911 - val_mean_absolute_error: 1.0581\n",
      "Epoch 105/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3097 - mean_squared_error: 10.0075 - mean_absolute_error: 1.3097 - val_loss: 1.0514 - val_mean_squared_error: 1.3764 - val_mean_absolute_error: 1.0514\n",
      "Epoch 106/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3691 - mean_squared_error: 10.6188 - mean_absolute_error: 1.3691 - val_loss: 1.0631 - val_mean_squared_error: 1.4021 - val_mean_absolute_error: 1.0631\n",
      "Epoch 107/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3814 - mean_squared_error: 10.7431 - mean_absolute_error: 1.3814 - val_loss: 1.0406 - val_mean_squared_error: 1.3526 - val_mean_absolute_error: 1.0406\n",
      "Epoch 108/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4131 - mean_squared_error: 11.0671 - mean_absolute_error: 1.4131 - val_loss: 1.0335 - val_mean_squared_error: 1.3372 - val_mean_absolute_error: 1.0335\n",
      "Epoch 109/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4742 - mean_squared_error: 11.6942 - mean_absolute_error: 1.4742 - val_loss: 1.0441 - val_mean_squared_error: 1.3604 - val_mean_absolute_error: 1.0441\n",
      "Epoch 110/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4255 - mean_squared_error: 11.1979 - mean_absolute_error: 1.4255 - val_loss: 1.0497 - val_mean_squared_error: 1.3725 - val_mean_absolute_error: 1.0497\n",
      "Epoch 111/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4621 - mean_squared_error: 11.5069 - mean_absolute_error: 1.4621 - val_loss: 1.0347 - val_mean_squared_error: 1.3399 - val_mean_absolute_error: 1.0347\n",
      "Epoch 112/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4616 - mean_squared_error: 11.5392 - mean_absolute_error: 1.4616 - val_loss: 1.0440 - val_mean_squared_error: 1.3600 - val_mean_absolute_error: 1.0440\n",
      "Epoch 113/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4523 - mean_squared_error: 11.3495 - mean_absolute_error: 1.4523 - val_loss: 1.0467 - val_mean_squared_error: 1.3659 - val_mean_absolute_error: 1.0467\n",
      "Epoch 114/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3591 - mean_squared_error: 10.4298 - mean_absolute_error: 1.3591 - val_loss: 1.0452 - val_mean_squared_error: 1.3628 - val_mean_absolute_error: 1.0452\n",
      "Epoch 115/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4939 - mean_squared_error: 11.7877 - mean_absolute_error: 1.4939 - val_loss: 1.0291 - val_mean_squared_error: 1.3277 - val_mean_absolute_error: 1.0291\n",
      "Epoch 116/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4183 - mean_squared_error: 10.9528 - mean_absolute_error: 1.4183 - val_loss: 1.0296 - val_mean_squared_error: 1.3288 - val_mean_absolute_error: 1.0296\n",
      "Epoch 117/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4276 - mean_squared_error: 10.9908 - mean_absolute_error: 1.4276 - val_loss: 1.0270 - val_mean_squared_error: 1.3233 - val_mean_absolute_error: 1.0270\n",
      "Epoch 118/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3904 - mean_squared_error: 10.6293 - mean_absolute_error: 1.3904 - val_loss: 1.0253 - val_mean_squared_error: 1.3196 - val_mean_absolute_error: 1.0253\n",
      "Epoch 119/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3940 - mean_squared_error: 10.6432 - mean_absolute_error: 1.3940 - val_loss: 1.0244 - val_mean_squared_error: 1.3175 - val_mean_absolute_error: 1.0244\n",
      "Epoch 120/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4244 - mean_squared_error: 10.8534 - mean_absolute_error: 1.4244 - val_loss: 1.0144 - val_mean_squared_error: 1.2961 - val_mean_absolute_error: 1.0144\n",
      "Epoch 121/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3211 - mean_squared_error: 9.8582 - mean_absolute_error: 1.3211 - val_loss: 1.0186 - val_mean_squared_error: 1.3052 - val_mean_absolute_error: 1.0186\n",
      "Epoch 122/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3596 - mean_squared_error: 10.2026 - mean_absolute_error: 1.3596 - val_loss: 1.0157 - val_mean_squared_error: 1.2990 - val_mean_absolute_error: 1.0157\n",
      "Epoch 123/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3577 - mean_squared_error: 10.1803 - mean_absolute_error: 1.3577 - val_loss: 1.0220 - val_mean_squared_error: 1.3125 - val_mean_absolute_error: 1.0220\n",
      "Epoch 124/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4409 - mean_squared_error: 11.0505 - mean_absolute_error: 1.4409 - val_loss: 1.0105 - val_mean_squared_error: 1.2878 - val_mean_absolute_error: 1.0105\n",
      "Epoch 125/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3426 - mean_squared_error: 10.0674 - mean_absolute_error: 1.3426 - val_loss: 1.0169 - val_mean_squared_error: 1.3015 - val_mean_absolute_error: 1.0169\n",
      "Epoch 126/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4800 - mean_squared_error: 11.3512 - mean_absolute_error: 1.4800 - val_loss: 0.9876 - val_mean_squared_error: 1.2398 - val_mean_absolute_error: 0.9876\n",
      "Epoch 127/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3577 - mean_squared_error: 10.0505 - mean_absolute_error: 1.3577 - val_loss: 1.0219 - val_mean_squared_error: 1.3123 - val_mean_absolute_error: 1.0219\n",
      "Epoch 128/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3418 - mean_squared_error: 9.9138 - mean_absolute_error: 1.3418 - val_loss: 0.9989 - val_mean_squared_error: 1.2635 - val_mean_absolute_error: 0.9989\n",
      "Epoch 129/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3804 - mean_squared_error: 10.2737 - mean_absolute_error: 1.3804 - val_loss: 0.9920 - val_mean_squared_error: 1.2489 - val_mean_absolute_error: 0.9920\n",
      "Epoch 130/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3569 - mean_squared_error: 9.9588 - mean_absolute_error: 1.3569 - val_loss: 1.0017 - val_mean_squared_error: 1.2693 - val_mean_absolute_error: 1.0017\n",
      "Epoch 131/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3071 - mean_squared_error: 9.5110 - mean_absolute_error: 1.3071 - val_loss: 1.0077 - val_mean_squared_error: 1.2821 - val_mean_absolute_error: 1.0077\n",
      "Epoch 132/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4585 - mean_squared_error: 10.9576 - mean_absolute_error: 1.4585 - val_loss: 0.9927 - val_mean_squared_error: 1.2506 - val_mean_absolute_error: 0.9927\n",
      "Epoch 133/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3708 - mean_squared_error: 10.0602 - mean_absolute_error: 1.3708 - val_loss: 0.9971 - val_mean_squared_error: 1.2598 - val_mean_absolute_error: 0.9971\n",
      "Epoch 134/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3532 - mean_squared_error: 9.9119 - mean_absolute_error: 1.3532 - val_loss: 0.9903 - val_mean_squared_error: 1.2455 - val_mean_absolute_error: 0.9903\n",
      "Epoch 135/200\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.3909 - mean_squared_error: 10.1775 - mean_absolute_error: 1.3909 - val_loss: 0.9954 - val_mean_squared_error: 1.2561 - val_mean_absolute_error: 0.9954\n",
      "Epoch 136/200\n",
      "43/49 [=========================>....] - ETA: 0s - loss: 1.4169 - mean_squared_error: 10.4742 - mean_absolute_error: 1.4169Restoring model weights from the end of the best epoch: 126.\n",
      "49/49 [==============================] - 0s 6ms/step - loss: 1.4209 - mean_squared_error: 10.5366 - mean_absolute_error: 1.4209 - val_loss: 0.9989 - val_mean_squared_error: 1.2634 - val_mean_absolute_error: 0.9989\n",
      "Epoch 136: early stopping\n",
      "######################\n",
      "Training Time: 41.64s\n",
      "batch size: 64\n",
      "n_layers: 1\n",
      "optimizer: <keras.optimizer_v2.adam.Adam object at 0x000001CBA5F46048>\n",
      "activation_function: relu\n",
      "loss_function mean_absolute_error\n",
      "Model: \"sequential_129\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " normalization_128 (Normaliz  (None, 4664)             9329      \n",
      " ation)                                                          \n",
      "                                                                 \n",
      " layer_normalization_481 (La  (None, 4664)             9328      \n",
      " yerNormalization)                                               \n",
      "                                                                 \n",
      " dense_620 (Dense)           (None, 1)                 4665      \n",
      "                                                                 \n",
      " layer_normalization_482 (La  (None, 1)                2         \n",
      " yerNormalization)                                               \n",
      "                                                                 \n",
      " activation_392 (Activation)  (None, 1)                0         \n",
      "                                                                 \n",
      " dropout_491 (Dropout)       (None, 1)                 0         \n",
      "                                                                 \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " dense_621 (Dense)           (None, 1)                 2         \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 23,326\n",
      "Trainable params: 13,997\n",
      "Non-trainable params: 9,329\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "29/29 [==============================] - 0s 4ms/step - loss: 0.9995 - mean_squared_error: 1.2733 - mean_absolute_error: 0.9995\n",
      "Test Loss: 0.9995102286338806 Test MSE: 1.273334264755249 Test MAE: 0.9995102286338806\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "trainer_reg = TrainerRegression(\n",
    "    units=2048,\n",
    "    activation_function=\"relu\", \n",
    "    optimizer=tf.keras.optimizers.Adam(learning_rate=1e-3), \n",
    "    n_layers=2,\n",
    "    loss_function='mean_absolute_error'\n",
    ")\n",
    "trainer_reg.train(x_train, y_train, \n",
    "                  batch_size=64,\n",
    "                  epochs=200)\n",
    "trainer_reg.print_summary()\n",
    "trainer_reg.plot_history()\n",
    "trainer_reg.evaluate(x_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Bonus"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test MSE: 0.07270271565619392 Test MAE: 0.16270201395509556\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.linear_model import Ridge\n",
    "\n",
    "def evaluate(x_test, y_test, y_pred):\n",
    "    print(\"Test MSE:\", np.sum((y_pred - y_test) ** 2) / len(y_test),\n",
    "          \"Test MAE:\", np.sum(np.abs(y_pred - y_test)) / len(y_test)\n",
    "    )\n",
    "    plt.title('Targets vs. Predictions')\n",
    "    plt.scatter(y_test, y_pred)\n",
    "    plt.xlabel('Target')\n",
    "    plt.ylabel('Prediction')\n",
    "    plt.show()\n",
    "\n",
    "clf = Ridge(alpha=1.0)\n",
    "clf.fit(x_train, y_train)\n",
    "evaluate(x_test, y_test, clf.predict(x_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Q3 - Dimensionality Reduction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## a)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "x_train: (50000, 32, 32, 3)\n",
      "y_train: (50000, 1)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "((50000, 1024), (50000, 10))"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(x_train, y_train), (x_test, y_test) = cifar10.load_data()\n",
    "print(\"x_train:\", x_train.shape)\n",
    "print(\"y_train:\", y_train.shape)\n",
    "\n",
    "def onehot(a):\n",
    "    a = a.flatten()\n",
    "    o = np.zeros((a.size, a.max() + 1))\n",
    "    o[np.arange(a.size), a] = 1\n",
    "    return o\n",
    "\n",
    "def preprocess(x, y):\n",
    "    assert x.shape[1:] == (32, 32, 3)\n",
    "    assert y.shape[1:] == (1, )\n",
    "    x = np.mean(x, axis=-1)  # grayscale\n",
    "    x = np.reshape(x, (-1, 32 * 32))  # flatten\n",
    "    x /= 255  # normalize\n",
    "    y = onehot(y)  # onehot labels\n",
    "    return x, y\n",
    "\n",
    "x_train_processed, y_train_processed = preprocess(x_train, y_train)\n",
    "x_test_processed, y_test_processed = preprocess(x_test, y_test)\n",
    "x_train_processed.shape, y_train_processed.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3sAAAEaCAYAAACo+k41AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOzdeXxU9b3/8feZmQTIQkhmgLBVJRgVEFkiIiKLRGtdqVdFREXRuqCibb2/KuB2FYtWRBGqVShSr9rYUuz1Vlwi4gKKIASVRZIKXiOBkAUhZM/5/v6YZMyQhQEmM8nk9Xw8xjnL95zzmfB9jHnnexbLGGMEAAAAAIgojnAXAAAAAAAIPsIeAAAAAEQgwh4AAAAARCDCHgAAAABEIMIeAAAAAEQgwh4AAAAARCDCHgCgXXvppZfkcrmCsq/jjz9ejz76aFD2FU6ZmZmyLEu7d+8OdykAgGNA2AOACHT99dfLsixZliWXy6XjjjtOt956qwoLC/3affrpp7rsssvUvXt3dezYUSkpKbrmmmu0YcOGBvt88skn5XQ69Zvf/CbgOkpLS/Xoo49q0KBBiomJUVJSks444ww9++yzKi0tPebPGS433XSTxo4d22D5unXr9Otf/7pFjz148GBdeeWVja7bs2ePoqKitGjRomM6xujRo5WXl6du3bod034AAOFF2AOACHX22WcrLy9PO3fu1Pz587Vs2TJdd911vvVLlizR2WefraioKL3yyivaunWrMjIydPzxx+uuu+5qsL8XX3xRM2bM0F/+8hdVVFQc9vj79+/XWWedpWeffVa333671qxZoy+++EL33HOPXn/9db377rtB/bytQdeuXRUbG9uix7j55pv1z3/+U3v37m2wbsmSJerUqZOuuuqqo95/ZWWloqOjlZycLIeDXxMAoE0zAICIM2XKFDN+/Hi/ZY8++qhxOBymtLTU/PDDD6ZDhw7mlltuaXT7oqIiv/mVK1eabt26maqqKtO/f3/zyiuvHLaGO+64w3Ts2NF8++23DdbZtm2Ki4ubrPXll1829f8X9eCDD5qUlBSTkZFh+vXrZzp16mQuvfRS8+OPP5ply5aZ1NRUExcXZ/7jP/7D7Nu3r9mfw6H7XrJkiXE6nX6fffLkyaZPnz6mY8eOJjU11Tz55JPGtm1fLZL8XkuWLDHGGHPccceZRx55xBhjzIwZM0xqamqDz37rrbeaM844wze/fv16c+6555rY2Fjj8XjML3/5S7Nz584mf64//vijiY2NNX/4wx8a/Ez79evn92/61FNPmUGDBpnY2FiTnJxsJk2aZPLy8nzr33vvPSPJvPXWW2bkyJGmQ4cO5rnnnvMtr2tbXV1tbrzxRtO3b1/TsWNH07dvXzNz5kxTUVHh29fMmTPNSSed5Pv3iI2NNePGjWvw7//555+b8847z8THx5u4uDgzfPhws27dOt/6t99+24wYMcJ07NjR9OzZ00ydOtUUFhY2+fMAADSNP9kBQDvRqVMn2bat6upqvf7666qoqNCsWbMabZuYmOg3/6c//UmTJ0+Wy+XSlClT9MILLzR7LNu29eqrr2ry5Mk64YQTGqy3LEtdunQ5ovrz8vK0dOlSLVu2TCtWrNDq1at1+eWXa9GiRXr99df11ltv6eOPP9Zjjz12RPs9VEVFhU499VS98cYb2rJli+6//349+OCDeumllyRJ99xzj66++mqdeeaZysvLU15eniZOnNhgP1OmTNH27dv16aef+pZVVlbq9ddf15QpUyRJW7Zs0ZgxY3TmmWdq/fr1WrlypZxOp84991yVl5c3Wl/nzp111VVX6cUXX/Rb/sEHHygnJ0c333yz3/KnnnpKX331lZYtW6Zvv/1WkydPbrDP3/72t7rvvvu0detWXXzxxQ3W27atnj176rXXXtPWrVs1d+5cvfjii3r88cf92uXm5urFF1/Ua6+9po8//liFhYW66aabfOu//PJLjRkzRh6PRytXrtSGDRt01113qaamRpL07rvv6pe//KWuueYaffXVV1q+fLmys7N1+eWXN/qzAAAcRrjTJgAg+A4d0dq8ebPp27evb0TptttuM507dw5oX/n5+SY6OtpkZWUZY4zZtWuXcblcZtu2bU1us2fPHiPJzJ0794hrNabxkT2n02n27t3rWzZt2jTjcDhMfn6+b9n06dPNsGHDjmjfh47sNWb69OkmPT3dN3/jjTeaMWPGNGhXf2TPGGPOOOMMc+utt/rmly1bZqKjo30jVVOmTDETJ07020d5ebnp1KmTWb58eZP1rF271kgyH374oW/ZVVdd5ffZG/P5558bSWb37t3GmJ9G9l599VW/doeO7DXmiSeeMCeffLJvfubMmcblcpmCggLfspdfftk4HA5TWVnpq3HIkCG+UdJDnXXWWWbmzJl+y/79738bSearr75q9rMBABpiZA8AItSqVasUFxenTp06aeDAgerbt69effVVSZIxJuD9LFmyRKeccopOO+00SVKPHj107rnnNhhZqq9u/5ZlHcMn8NerVy95PB7ffHJyspKTk9W1a1e/Zfn5+cd0HNu2NWfOHA0ePFgej0dxcXF6/vnn9d133x3xvq677jplZGSosrJSkvTyyy/r4osvVlJSkiTvDV2WL1+uuLg438vtdqu8vFzZ2dlN7nf48OEaMmSI79+gsLBQy5cvbzCqt3LlSp133nnq06eP4uPjfTeVOfSzDB8+/LCf5fnnn9fpp5+ubt26KS4uTvfff3+D/fTp00dut9s336tXL9m27bu+8IsvvlB6enqT/WL9+vV68skn/X4egwYNkqRmfx4AgMYR9gAgQp1xxhnKysrS1q1bVVZWpvfee099+/aVJJ100knav3+/cnNzm92HMUaLFi3Sl19+KZfL5Xu9/fbbWrp0qS/EHKpr165KTEzU5s2bD1unw+FoED6rqqoatIuKivKbtyyr0WW2bR/xvuubO3eufv/73+vOO+/Ue++9p6ysLN10001NftbmXHXVVTp48KDefPNNFRUV6a233vK7SY5t27r22muVlZXl99q+fbvf6Y+Nufnmm/X3v/9dxcXFWrp0qaKiojRp0iTf+h07dujCCy9USkqK/vrXv2r9+vVavny5JDX4LIe7qcxrr72mu+66S5MnT9aKFSu0ceNGzZw5s8F+oqOj/ebrQl39f5Pm/gBg27ZmzpzZ4OeRnZ2tc889t9kaAQANBefBQgCAVqdTp07q169fo+uuuOIK3XvvvXr00Uf1/PPPN1hfXFysxMRErVy5Ut9++61Wr16t+Ph433rbtjV69Gj94x//aPTOjw6HQ1dffbUWL16smTNnNrhuzxij/fv3KyEhQd26dfO7rk1So49+OBpHs++PPvpI559/vm688UbfskNHlaKjo33XmTUnKSlJF110kf7yl79oz549SkhI0C9+8Qvf+rS0NH355ZdKSUk54lHQyZMn6z//8z/18ssva9GiRZo0aZLfv9Hnn3+uiooKPfPMM74Qtnbt2iM6Rp2PPvpIaWlpuvvuu33LduzYccT7GTZsmN577z3NmTOn0c87bNgwbd68ucl+CwA4MozsAUA71KtXLy1YsEAvvviirrrqKr3//vvauXOnNmzYoAcffFCXXnqpJO+NWepuIDJw4EDfa9CgQbrkkkuavVHL7NmzdeKJJ2rEiBF64YUXtGnTJu3YsUPLly/XmDFj9MEHH0iS0tPTtW3bNi1YsED//ve/9eKLL+r1118Pyuc8mn2fdNJJWrVqlT744ANt375ds2bNahCSTjjhBG3btk2bN29WQUFBs4+imDJlilasWKE//vGPmjRpkt9o5IwZM7R161Zdc801+vzzz7Vjxw598MEHuuuuu/Ttt982W2d8fLwmTZqkhx56SFu3bm1wCmdqaqqMMZo7d67v5360D3w/6aSTlJWVpTfffFM5OTmaN2+e/vnPfx7xfn73u99py5Ytuvbaa/XFF18oJydHGRkZvp/vI488omXLlumee+5RVlaWcnJytGLFCt1www1HNbIKAO0dYQ8A2qmbbrpJH374ocrLyzVp0iSddNJJuvzyy7Vjxw7Nnz9f+fn5euONN5p8gPfEiRO1atWqJq+lSkhI0Keffqpp06Zp/vz5GjFihIYOHao5c+Zo4sSJ+vnPfy7JG8geffRR/f73v9dpp52mlStX6oEHHgjKZzyafd9///0aM2aMLr30Up155pkqLi7W9OnT/drceOONOv300zVy5Eh17dpVr732WpP7+8UvfqEuXbpo8+bNfqdwStIpp5yiNWvWqKSkRD//+c/Vv39//epXv1JZWVlAdyu95ZZbVFxcrMGDBystLc1v3ZAhQ/TMM89o4cKF6t+/v+bNm6enn376sPtszLRp0zRp0iRNmTJFw4YN0xdffHFU/0aDBw/WqlWrlJeXp9GjR2vIkCF6+umn5XJ5TzRKT09XZmamNmzYoFGjRum0007Tb3/7W3Xp0kVOp/OoageA9swyR3KVPgAAAACgTWBkDwAAAAAiEGEPAAAAACIQYQ8AAAAAIhBhDwAAAAAiEGEPAAAAACJQm3+o+q5du8JdQgMej0cFBQXhLgMRhD6FYKNPIZjoTwg2+hSCLZL7VM+ePZtcx8geAAAAAEQgwh4AAAAARCDCHgAAAABEIMIeAAAAAEQgwh4AAAAARKCQ3I3zj3/8ozZs2KCEhATNnTu3wXpjjJYsWaKNGzeqQ4cOmjZtmvr27RuK0gAAAAAgIoVkZG/s2LGaMWNGk+s3btyo3bt3a/78+br55pu1aNGiUJQFAAAAABErJCN7/fv3V35+fpPr169fr9GjR8uyLKWmpurgwYMqLi5WYmJiKMoDAACAvGdbyRhJRjLyn/b+x7vs0HnfdCPzqttnI/N+x/lpvsYyMkWF9Y5dbztf26aOXW+/DWquf/zmtqt/vEO3UwA1HPJ5fTUcSe2HHvOQ+frtmvmZm/r7qr+gfi2N7uOQmuovb+pn0dw2je2/QU2N7D+gmprZT+3yg7GxsktKAqyzif0nuuUYfb7aklbxUPWioiJ5PB7fvNvtVlFRUaNhLzMzU5mZmZKkOXPm+G3XWrhcrlZZF9ou+hSCjT7VdhljJNuW7BrJtmVq31VTN10j2cb7boyMbdf+kmn7lpv687Xvxtje/dRb/tOxbP92vn15pystS7HV1X7L/dqZ2hpsu7Ye/+XGDuDY9T9LveW+dr7PZPu384WJujam9ne3er/wH7Je9dabunm/fdVtV+8X6ibXyzftbdPYsf3313TgasH91e2rlYjMR18jnEoO36R5lqWo1AFKuuyaYJQTMq0i7Pm+SOuxLKvRtunp6UpPT/fNFxS0vq8Dj8fTKutC20WfQrC1pT5l7BqpxpZqqr0Bpqbey67xLm+wvvqQNt53U11dG4ya2lddWLLrvWr8Q4dt17bzn/eFpca2q2tb08S+TDPbNVZDJLOs2pfD++6wJMtZ+15vuWVJDod33nHIcstRu846ZJ+WJEuyVG+6/jo1nFZtDTq0jeOn/fjtO5D9SJYaOe4hbX6qr/50Xft6x2/Q3r+N1WibQ9ofrs2hn8nv51hvXoduX38flm+2wbzfcaS4uHiVlJQc8vM59PM2dyz91L5ejT9NNrGf+vXUb9dc3b7jBFbDEdV+zDXUbyf/Ouovb2offtscpq7Gtjmi/TdW6xHsv9n91P1/r7CZf/OG+z80j9hqndmjZ8+eTa5rFWHP7Xb7/eAKCws5hRMAgsQY81MAqq6SqqtVY6pl9uZLVdVSjXeZ91VV267aG4yqq/yWqaZaqqo3X3+beiHL+AWpegHr0GB16PrGAlgjfxBscXVBwuGsfa8LFQ7J6fxp2vdyNj/vdErOaP/9OByyGmx36LaHrLdq5511+2mira/+2nb1wpF1aBjyhaVGlvvCVvPLE5OSVLzvR7/j+IWxQ49z6PJ6+2vqj71oX2I8HpW2wl+q0XZZUdGyoqLCXUbItYqwl5aWprfffltnnXWWsrOzFRMTQ9gD0KYZY7whqKrSG4783iv91pn666oba1/l28bUhrX6wa3REOYLcLXvhwjKr1CuKMnl8r6cUbWBpvblqPfucnl/uXdFSdEd/dpZTtdPbf22dXkDTXPrHbXrnU5ZR7Kd86ftGiyzvMGMwHFkXB6PrA78Yg4ArU1Iwt7TTz+tLVu26MCBA7r11lt15ZVXqrr2l4/zzjtPQ4YM0YYNGzR9+nRFR0dr2rRpoSgLQDtjjPGGpsoK76ui4qfp2pepbLjM27a83vp6+6iqajygVVcde8GuKCkqWoqqfXe6vNN1Icvpkjp28s1bztrg5VtfL4zV38blUnyXJB0oL5OcUbIaWd8gyEXV219dwCIQAQDQqoUk7N19993NrrcsSzfddFMoSgHQBhhjvOGqolwqL/O+Krzvpm5Z7bx3uvyndU2EON/rSFmWFN2h4SsqWuoUK3WOklU/kEVFNwxpde+u2rbRh7ZrpK3TJcvRck/H6eTx6CCnSAEAENFaxWmcACKDqa6SykqlsoPe99KDUtlBmbplpQelsnrBrV5Q8wt2leWBX6flivKObnXo6H2vC2MJibLqpjscGtY6egNXdAdZDdbVbdPR++6KYgQLAAC0SYQ9AD6muloqPSAdLJEOet9NbWCrH95UVuq/vKx2urLy8Afp0MkbyuoHtIREWb5lnaSOHf3WW41t08E7bbn4GgMAAGgMvyUBEcjYtjewleyvDW0HZGrDm0oO+AKdKdlfL9gd8I6qNSe69tTFTjG177GykrpKMbE/La+dtupN+5Z37OS9+yAAAABaHGEPaAOMMTKlJdL+H6UD3pc58NO03/z+fd5A19SzuCyHFBsnxcZ73xMSZfX8mRQXX295vKy69fUCm+Vqf7csBgAAaKsIe0AYmepqbzj7sUj6sUhmX3HtdLHMPu+79hcrv2R/o7fPl+QNYvEJUucEqWsPWSknS3G183Gda0NbbXCLi5c6xrTojT8AAADQOhD2gBZiSg9KRXulor0yRXul4iLpx0KZH4ululBXsr/hjUgsS+rcRUpIlBKSZPU5QZ2691CZK1qK7ywrvos33MXXhrl2+IBQAAAAHB5hDzgKprpKKi6Uigu8Qa5wr1RUO133OvT6N4dD6pzoDXHurrL6nuSd7pIoK8EtdaldF9/F+4DoeuI9HlVwm3wAAAAcAcIe0ARTelDau1vamyezd7e0d7dMfp53WXFBwxG5uM5SUlepW09Zp5wmJXm8Ny+pe3VO4OYkAAAACBnCHto1U1oi5eXK7PlBqg1ypjbgqeSAf+P4BKlbD1mpAyRPsnd0ri7IJXq8z2sDAAAAWgnCHiKeMcZ7yuXuXJm8XGn397Xvud4boNSxHJK7q9Q1WdbQs6RuybK6Jktde0hdu8vqGBO+DwEAAAAcIcIeIoopPSjl7pTJ3VH7vlPK+97/+rlOMVJyb1kDhko9esvq0Ufq3kvydOcB3QAAAIgY/GaLNsnYtlSwR8rdIfN9bbj7fodUmP9To9h4qffxss48R+rRR1ZyL6lHH+9z5SwrfMUDAAAAIUDYQ6vnOw1zZ7bMzu0yO3OknTlS2UFvA8shde8p64RU6ezzZPU5Qep9gpToJtQBAACg3SLsodUxVZXSju0y2VtkdmyXdmb/dG2d0yn1Ol7W6WdLx6XI6tNX6vkzbo4CAAAAHIKwh7Az5WXSt9tktm+Wyd4sfbtdqq7yrkzuLeuUwdLxJ8o64USpzwmyoqLDWzAAAADQBhD2EHKmskLK3iKzJcsb7r7LkWzb+9Dx4/rJOuci7+MN+p0iKzY+3OUCAAAAbRJhDy3OGCPt+l5m8waZzRul7M1SVaXkckknpMo6/3JvuEs5iccbAAAAAEFC2EOLMJUV0tZNMps+l/l6g1Rc4F3Ro4+sMefL6j9ESh3ItXYAAABACyHsIWhMyX6ZL9fJZK2VNm+UKiu8z7TrP1jWgKtk9R8iy9013GUCAAAA7QJhD8fElJfKZK2VWfuhtCXLe+1dokfWWeNlDT7DO3rnigp3mQAAAEC7E7Kwl5WVpSVLlsi2bY0fP14TJkzwW793714999xz2r9/v+Li4nTnnXfK7XaHqjwcAVNdJW3eKLP2Q5lNa6XKSsndTdbPfylr2FnSz1J4vh0AAAAQZiEJe7Zta/HixZo1a5bcbrfuu+8+paWlqXfv3r42L7/8skaPHq2xY8fq66+/1quvvqo777wzFOUhQCYvV+bjd2Q+XSmVHJDi4mWNHC/rjDFS35NlORzhLhEAAABArZCEvZycHCUnJ6t79+6SpJEjR2rdunV+YS83N1dTpkyRJA0YMEB/+MMfQlEaDsNUVclsWCPz0TvS9q+9DzU/7Qw5Ro6XBgzmFE0AAACglQpJ2CsqKvI7JdPtdis7O9uvzXHHHae1a9fqggsu0Oeff66ysjIdOHBA8fH+z1nLzMxUZmamJGnOnDnyeDwt/wGOkMvlapV1HYma4kKVrVim0nfekNm/T87uPdXpmlvV8ZwL5Uzk9NpQi4Q+hdaFPoVgoj8h2OhTCLb22qdCEvaMMQ2WHXpN17XXXqs///nPWrVqlU455RQlJSXJ6XQ22C49PV3p6em++YKCguAXfIw8Hk+rrCsQ5of/k3nvDZm1q6SaGum04XKMvUDmlNNU5nCorMZIbfSztWVtuU+hdaJPIZjoTwg2+hSCLZL7VM+ePZtcF5Kw53a7VVhY6JsvLCxUYmKiX5ukpCTdc889kqTy8nKtXbtWMTE8YDtUzHf/lv3ma9Kmz6WoaFmjzpWVfqms7k13HgAAAACtV0jCXkpKivLy8pSfn6+kpCStWbNG06dP92tTdxdOh8Oh5cuXa9y4caEord0zuTtk//M1KeszKSZO1iVXyxp7gaz4zuEuDQAAAMAxCEnYczqdmjp1qmbPni3btjVu3Dj16dNHGRkZSklJUVpamrZs2aJXX31VlmXplFNO0Y033hiK0tots69QZtlfZD77QOoUI+viSbLSL5EVExvu0gAAAAAEgWUau6CuDdm1a1e4S2igNZ8TbKqqZDL/R+Zfr0s1VbLGXyLrF/8hKzb+8BsjbFpzn0LbRJ9CMNGfEGz0KQRbJPepsF+zh9bBfLlOdsYiKT/Pe+OVK6fK6sY1eQAAAEAkIuy1A+bAjzKvvSCz7mMpubccdz0ka+DQcJcFAAAAoAUR9iKYMUZm3ccyr70glZV6b77yi//gQegAAABAO3BEYa+goEBFRUVKTU1tqXoQJKa0RPZfFkhfrJGOP1GO66fL6nVcuMsCAAAAECIBhb2CggI988wz2rlzpyTp5Zdf1meffaasrCzdeuutLVkfjoLZkS37hSek4gJZl02Rdd4EWY08oB4AAABA5HIE0uiFF17QkCFDtHTpUrlc3nw4aNAgffnlly1aHI6MMUb2yv+V/fjvJNuW4z9/L8cv/oOgBwAAALRDAY3s5eTk6N5775XD8VM2jImJUWlpaYsVhiNjqqtkXv2TzMfveu+0ecNdPE4BAAAAaMcCCnsJCQnavXu33zMccnNz5fF4WqwwBM6UHpS9cLa0/WtZF1wp69KrZTkCGrQFAAAAEKECCnsXX3yxHn/8cU2YMEG2beuTTz7R8uXLNWHChJauD4dh9u+T/cxD0g/fybrxN3KMGBvukgAAAAC0AgGFvXPOOUdxcXF6//335Xa79dFHH2nixIkaPnx4S9eHZpjCvbLnPSAV75XjjlmyBg4Ld0kAAAAAWomAH70wfPhwwl0rYor2yv7DfVLpQTnu/i9ZJ/YPd0kAAAAAWpGALuz685//rG+++cZv2TfffKOXXnqpJWrCYZj9xbKfekAqLZHjt48Q9AAAAAA0EFDYW716tVJSUvyW9e3bV5988kmLFIWmmYMHvEGvuECO6Q/IOq5fuEsCAAAA0AoFFPYsy5Jt237LbNuWMaZFikLjTEWF7Pn/Je35QY7bZ8rqx4geAAAAgMYFFPZOPvlk/fWvf/UFPtu29be//U0nn3xyixaHnxjblv3nedKO7XL86j9l9R8c7pIAAAAAtGIB3aDlhhtu0Jw5c3TLLbfI4/GooKBAiYmJ+t3vftfS9aGWeet1acMaWVfeKGvomeEuBwAAAEArF1DYc7vdevzxx5WTk6PCwkK53W7169dPDh7cHRJm25cy//NXWSPGykq/JNzlAAAAAGgDAn70gsPhUGpqqt+1e7ZtE/hamNlfLHvRXKl7D1mTb5NlWeEuCQAAAEAbEFDY+/bbb7V48WL93//9nyorK/3WZWRktEhhkIxdI3vRU7XP0ntYVsdO4S4JAAAAQBsRUNhbuHChhg0bpttuu00dOnRo6ZpQy6z8l7R1k6zr7pDV+/hwlwMAAACgDQko7BUUFGjSpEnHdAphVlaWlixZItu2NX78eE2YMKHBMRYuXKiDBw/Ktm1dffXVGjp06FEfr60zRXtl3nhFGjhM1qhzw10OAAAAgDYmoAvuTj/9dG3atOmoD2LbthYvXqwZM2Zo3rx5Wr16tXJzc/3aLFu2TGeeeaaeeOIJ3X333Vq8ePFRHy8S2H99UTI1clx9C9fpAQAAADhiAY3sVVVV6cknn9TJJ5+sLl26+K274447Drt9Tk6OkpOT1b17d0nSyJEjtW7dOvXu3dvXxrIslZaWSpJKS0uVmJgY8IeINGbbl9LGz2RNuEZW1+RwlwMAAACgDQoo7PXu3dsvmB2poqIiud1u37zb7VZ2drZfmyuuuEKPPvqo3n77bVVUVOj+++9vdF+ZmZnKzMyUJM2ZM0cej+eo62opLpfrqOsyNTUq+sdSqWt3ea66URbXSELH1qeAxtCnEEz0JwQbfQrB1l77VEBh74orrjimgxhjGiw79NTE1atXa+zYsbr44ou1fft2Pfvss5o7d26DRzukp6crPT3dN19QUHBMtbWEugfPHw37s1UyO7Jl3fRbFR44IB04EOTq0BYdS58CGkOfQjDRnxBs9CkEWyT3qZ49eza5LuDn7FVXV2vXrl3av3+/3/KBAwcedlu3263CwkLffGFhYYPTNFeuXKkZM2ZIklJTU1VVVaUDBw4oISEh0BLbPGPXyPxvhtTrOFmnnx3ucgAAAAC0YQGFvW3btow0S0cAACAASURBVOmpp55SVVWVysrK1KlTJ5WXl8vtdmvBggWH3T4lJUV5eXnKz89XUlKS1qxZo+nTp/u18Xg8+vrrrzV27Fjl5uaqqqpKnTt3PrpP1UaZzz+W9vwgx633yuJh9QAAAACOQUBhb+nSpbrkkkt00UUX6YYbbtCSJUv097//XdHR0QEdxOl0aurUqZo9e7Zs29a4cePUp08fZWRkKCUlRWlpabruuuv0pz/9Sf/6178kSdOmTWtXd6GsP6qnISPCXQ4AAACANi6gsLdr1y5dcMEFfssmTJig22+/XZdccklABxo6dGiD5+ZNnDjRN927d2898sgjAe0rEjGqBwAAACCYAkoVMTExKisrkyR16dJFubm5KikpUXl5eYsW114wqgcAAAAg2AIa2TvjjDO0ceNGjRo1Suecc44efvhhOZ1OnXnmmS1dX/uQtbZ2VO93jOoBAAAACIqAwt7111/vm7744ovVr18/lZeX67TTTmuputoV+8N3pESPNJhRPQAAAADBEfCjF+o75ZRTgl1Hu2X27pa2bJR18SRZTme4ywEAAAAQIZoMe7Nnz9bMmTMlSQ888ECTd8Z8+OGHW6aydsJ8/I5kOWSNOjfcpQAAAACIIE2GvTFjxvimzznnnJAU094Yu0bm01XSqcNkJXnCXQ4AAACACNJk2Bs1apQkybZt7dmzR5dddpmioqJCVli7kL1F2lco68qp4a4EAAAAQIQ57K0fHQ6H3nnnHTm5nizozNoPpQ4dZQ0aHu5SAAAAAESYgO7zP2bMGL333nstXUu7YqqqZL5YLWvICFkdOoS7HAAAAAARJqC7cebk5Ojtt9/W//zP/8jtdvvdrIUbtBylb76USg/KOv3scFcCAAAAIAIFFPbGjx+v8ePHt3Qt7YrZnCW5oqSTB4W7FAAAAAARKKCwN3bs2BYuo/0xmzdIqQNkRXMKJwAAAIDgC/ih6vv27VNOTo4OHDggY4xvOY9lOHKmqEDK+17WWenhLgUAAABAhAoo7H3++ed69tln1aNHD33//ffq06ePvv/+e5188smEvaNgtmZJkqwBQ8JcCQAAAIBIFVDYy8jI0LRp03TmmWfqhhtu0BNPPKEPPvhA33//fUvXF5k2b5QSEqVex4W7EgAAAAARKqBHLxQUFOjMM8/0WzZmzBh99NFHLVJUpDM5W2WlDvS7qykAAAAABFNAYa9z587at2+fJKlr167avn279uzZI9u2W7S4SGQK90rFBVK/U8JdCgAAAIAIFvCjF7Zt26YRI0bowgsv1MMPPyzLsnTRRRe1dH0Rx+RskSRZhD0AAAAALSigsDdhwgTf9JgxYzRgwACVl5erd+/eLVZYxPr3VqlDR6nX8eGuBAAAAEAEa/Y0zqeeekobNmxocLqmx+Mh6B0lk7NV6nuSLKcz3KUAAAAAiGDNjuwlJSXp+eeflzFGo0aN0tixY3XccUd3B8msrCwtWbJEtm1r/PjxfqOFkvTSSy9p8+bNkqTKykr9+OOPeumll47qWK2VKS+Vcr+TdeGV4S4FAAAAQIRrNuxdf/31uu6665SVlaWPP/5Ys2bNUnJyssaMGaNRo0apS5cuAR3Etm0tXrxYs2bNktvt1n333ae0tDS/0cHrr7/eN71ixQrt2LHj6D5Ra/btdsnYXK8HAAAAoMUd9po9h8OhoUOHaujQoSotLdVnn32mjz/+WK+99ppOPfVU3XvvvYc9SE5OjpKTk9W9e3dJ0siRI7Vu3bomTwVdvXq1rrwy8ka/TM5WybKkE1LDXQoAAACACBfQDVrqxMTEaMiQISopKdGePXu0devWgLYrKiqS2+32zbvdbmVnZzfadu/evcrPz9fAgQMbXZ+ZmanMzExJ0pw5c+TxeI7kI4SEy+VqtK7i/8uRfVyK3D/jYeo4Mk31KeBo0acQTPQnBBt9CsHWXvtUQGGvsrJSn3/+uT788ENt3rxZp5xyiiZOnKgRI0YEdBBjTINlTT1QfPXq1RoxYoQcjsbvHZOenq709HTffEFBQUA1hJLH42lQlzFG9vbNsoaPbpU1o3VrrE8Bx4I+hWCiPyHY6FMItkjuUz179mxyXbNhb/Pmzfrwww+1du1aJSYmavTo0brllluOOBW73W4VFhb65gsLC5WYmNho2zVr1ujGG288ov23CUV7pbJSqU/fcFcCAAAAoB1oNuw9+eSTGjlypGbOnKnU1KO/ziwlJUV5eXnKz89XUlKS1qxZo+nTpzdot2vXLh08ePCYjtVq/fCdJMnq9bMwFwIAAACgPWg27L3wwguKioo65oM4nU5NnTpVs2fPlm3bGjdunPr06aOMjAylpKQoLS1NkvTJJ59o5MiRTZ7i2ZaZH/7PO9GTsAcAAACg5TUb9oIR9OrU3dGzvokTJ/rNR+IdOH12fSclemTFxIW7EgAAAADtQON3QUHQmR++k3pxF04AAAAAoUHYCwFTUyPl5XK9HgAAAICQOaKwV1BQoO3bt7dULZErP0+qrpJ6MrIHAAAAIDQCes5eQUGBnnnmGe3cuVOS9PLLL+uzzz5TVlaWbr311pasLzLsqrsTJ2EPAAAAQGgENLL3wgsvaMiQIVq6dKlcLm8+HDRokL788ssWLS5SmD27vBPJvcJbCAAAAIB2I6Cwl5OTowkTJsjh+Kl5TEyMSktLW6ywiFJcIMXGy+rQMdyVAAAAAGgnAgp7CQkJ2r17t9+y3NxceTyeFikq0pjiQinRHe4yAAAAALQjAV2zd/HFF+vxxx/XhAkTZNu2PvnkEy1fvlwTJkxo6foiQ9FeKZFgDAAAACB0Agp755xzjuLi4vT+++/L7Xbro48+0sSJEzV8+PCWri8yFBfI6ntSuKsAAAAA0I4EFPZs29bw4cMJd0fBVFZIJQcY2QMAAAAQUgFds/erX/1KixYt0rZt21q6nshTXOh9J+wBAAAACKGARvZmzZql1atX65lnnpHD4dBZZ52lUaNG6Wc/+1lL19f2Fe2VJFlJhD0AAAAAoRNQ2DvhhBN0wgkn6JprrtGWLVv0ySef6L/+67/UpUsXPfnkky1dY5tmGNkDAAAAEAYBncZZX8+ePdW7d2+53W7t3bu3JWqKLMUF3ncevQAAAAAghAIa2Tt48KDWrl2rTz75RNnZ2Ro0aJAuvfRSpaWltXR9bV9RgRTXWVZ0h3BXAgAAAKAdCSjs3XLLLTrppJM0atQo3XPPPYqJiWnpuiKGKS5gVA8AAABAyAUU9p599lklJia2dC2RqbhAcncLdxUAAAAA2pkmw96WLVvUv39/SdIPP/ygH374odF2AwcObJnKIsW+Qln9Tgl3FQAAAADamSbD3uLFizV37lxJ0nPPPddoG8uytGDBgpapLAKYmhrvA9Xju4S7FAAAAADtTJNhry7oSdLChQuP+UBZWVlasmSJbNvW+PHjNWHChAZt1qxZo7/97W+yLEvHHXec7rrrrmM+bliV7Pe+d04Ibx0AAAAA2p2AHr3wxBNPNLo80Gfs2batxYsXa8aMGZo3b55Wr16t3NxcvzZ5eXl644039Mgjj+ipp57S9ddfH9C+W7UD+yRJFiN7AAAAAEIsoLC3efPmI1p+qJycHCUnJ6t79+5yuVwaOXKk1q1b59fm/fff189//nPFxcVJkhISImA0bP+P3vf4zuGtAwAAAEC70+zdODMyMiRJ1dXVvuk6e/bsUdeuXQM6SFFRkdzunx4/4Ha7lZ2d7ddm165dkqT7779ftm3riiuu0ODBgwPaf2tlDtSFPUb2AAAAAIRWs2GvsLBQkvc0zLrpOh6PR1deeWVABzHGNFhmWZbfvG3bysvL04MPPqiioiI98MADmjt3rmJjY/3aZWZmKjMzU5I0Z84ceTyegGoIJZfLJY/Ho4N2tUokuY/vKwejezgGdX0KCBb6FIKJ/oRgo08h2Nprn2o27E2bNk2SlJqaqvT09KM+iNvt9guLhYWFDZ7bl5SUpNTUVLlcLnXr1k09e/ZUXl6e+vXr59cuPT3dr5aCgoKjrquleDweFRQUyN69S3I6VVhWLquiMtxloQ2r61NAsNCnEEz0JwQbfQrBFsl9qmfPnk2uC+iavbpwVVZWpvz8fO3Zs8f3CkRKSory8vKUn5+v6upqrVmzRmlpaX5thg8frq+//lqStH//fuXl5al79+4B7b/VOvCjFJcgyxHQjxkAAAAAgqbZkb06ubm5mj9/vr777rsG6w69lq8xTqdTU6dO1ezZs2XbtsaNG6c+ffooIyNDKSkpSktL02mnnaZNmzbp17/+tRwOh6655hrFx8cf+SdqRcz+fVJ8BNxoBgAAAECbE1DYW7RokQYMGKAHH3xQd9xxhxYuXKhXX31VqampAR9o6NChGjp0qN+yiRMn+qYty9KUKVM0ZcqUgPfZ6h34kWfsAQAAAAiLgM4v/O677zR58mTFxsbKGKOYmBhdc801AY3qtWsHfpQVR9gDAAAAEHoBhb2oqCjV1NRIkuLj41VQUCBjjEpKSlq0uDZvPyN7AAAAAMIjoNM4Tz75ZH366acaO3asRowYoccee0xRUVEaMGBAS9fXZpmKCqmijGv2AAAAAIRFQGHvN7/5jW960qRJ6tOnj8rLyzV69OgWK6zNK6l7oDphDwAAAEDoBRT26nM4HIS8QBzwhj2rc5cwFwIAAACgPWoy7D377LOyLOuwO7jjjjuCWlDEOMDIHgAAAIDwaTLsJScnh7KOiGP2E/YAAAAAhE+TYe+KK64IZR2R5+AB73ts234wPAAAAIC2KaBr9r7++usm1w0cODBoxUSU0hLJckgdO4W7EgAAAADtUEBh77nnnvOb379/v6qrq+V2u7VgwYIWKazNKy2ROsXIcgT0KEMAAAAACKqAwt7ChQv95m3b1rJly9SpE6NWTTp4UIqNC3cVAAAAANqpoxp2cjgcuuyyy/TPf/4z2PVEDFN2UIoh7AEAAAAIj6M+x/DLL7+Ug1MUm1ZaIsXEhrsKAAAAAO1UQKdx3nbbbX7zlZWVqqys1E033dQiRUWEgyWyurjDXQUAAACAdiqgsHfnnXf6zXfo0EE9evRQTExMixQVERjZAwAAABBGAYW9/v37t3QdkaeUa/YAAAAAhE9AYa+0tFRvvfWWdu7cqfLycr91s2bNapHC2jJTUSFVV3E3TgAAAABhE1DYe+qpp2TbtoYPH67o6OiWrqnNsw/u904wsgcAAAAgTAIKe9nZ2Vq8eLFcroCat3um9KB3oiPPIQQAAAAQHgE9O+Hkk0/WDz/80NK1RAxTWSFJsjp0CHMlAAAAANqrgIbqpk2bpt///vfq16+funTp4rfu8ssvD+hAWVlZWrJkiWzb1vjx4zVhwgS/9atWrdLLL7+spKQkSdL555+v8ePHB7Tv1sZUVnonogh7AAAAAMIjoLD32muvqbCwUF27dlVZWZlvuWVZAR3Etm0tXrxYs2bNktvt1n333ae0tDT17t3br93IkSN14403HkH5rZOpqL2JTRTXNwIAAAAIj4DC3po1a/TMM88oMTHxqA6Sk5Oj5ORkde/eXZI31K1bt65B2IsYtadxipvZAAAAAAiTgMJe9+7d5XQ6j/ogRUVFcrvdvnm3263s7OwG7dauXautW7eqR48emjJlijweT4M2mZmZyszMlCTNmTOn0TbhVpldJUlK7NZdrlZYH9oel8vVKvs62i76FIKJ/oRgo08h2Nprnwoo7J199tl64okndP755ze4Zm/gwIGH3d4Y02DZoaeADhs2TGeddZaioqL07rvvauHChXrwwQcbbJeenq709HTffEFBQSAfIaRia091LT5YKqsV1oe2x+PxtMq+jraLPoVgoj8h2OhTCLZI7lM9e/Zscl1AYe+dd96R5L12rz7LsrRgwYLDbu92u1VYWOibLywsbHBKaHx8vG86PT1dr7zySiCltUrGdxonN2gBAAAAEB4Bhb2FCxce00FSUlKUl5en/Px8JSUlac2aNZo+fbpfm+LiYl8AXL9+fZu+no8btAAAAAAIt5A8Jd3pdGrq1KmaPXu2bNvWuHHj1KdPH2VkZCglJUVpaWlasWKF1q9fL6fTqbi4OE2bNi0UpbUIww1aAAAAAIRZQGHvtttua3Ldc889F9CBhg4dqqFDh/otmzhxom/66quv1tVXXx3Qvlq9ygrJckjOkGRpAAAAAGggoDRy5513+s0XFxfrrbfe0llnndUiRbV1prJCio4O+DmEAAAAABBsAYW9/v37N1g2YMAAzZ49WxdccEHQi2rrvGGPm7MAAAAACB/H0W7ocrmUn58fzFoihqmo4OYsAAAAAMIqoJG9jIwMv/mKigpt3LhRQ4YMaZGi2rq60zgBAAAAIFwCCnv1n5EnSR06dNBFF12k0aNHt0hRbZ2pZGQPAAAAQHgFFPba8mMQwoJr9gAAAACEWbPX7G3btk3//d//3ei6V155Rdu3b2+Roto6U1HOyB4AAACAsGo27C1fvrzRO3FK3rtx/uMf/2iRoto67sYJAAAAINyaDXs7d+7U4MGDG1136qmnaseOHS1SVFtnKipkMbIHAAAAIIyaDXtlZWWqrq5udF1NTY3KyspapKi2jrtxAgAAAAi3ZsNer169tGnTpkbXbdq0Sb169WqRotq86irJFRXuKgAAAAC0Y82GvQsvvFAvvPCC1q5dK9u2JUm2bWvt2rV68cUXdeGFF4akyLbGVFdLzoBudAoAAAAALaLZRDJq1Cjt27dPCxcuVFVVlTp37qz9+/crOjpaV1xxhUaNGhWqOtuWmmrJ6Qx3FQAAAADascMOP1100UU655xztH37dpWUlCguLk6pqamKiYkJRX1tkqmulsXIHgAAAIAwCiiRxMTENHlXTjSCkT0AAAAAYdbsNXs4csYYqaaGa/YAAAAAhBVhL9hqarzvjOwBAAAACCPCXrD5wh4jewAAAADCh7AXbDW1D6FnZA8AAABAGBH2go2RPQAAAACtQMjCXlZWlu666y7deeedeuONN5ps99lnn+nKK6/Uv//971CVFlyM7AEAAABoBUIS9mzb1uLFizVjxgzNmzdPq1evVm5uboN2ZWVlWrFihU488cRQlNUyuEELAAAAgFYgJGEvJydHycnJ6t69u1wul0aOHKl169Y1aJeRkaFLLrlEUVFRoSirZfhG9jiNEwAAAED4hCSRFBUVye12++bdbreys7P92uzYsUMFBQUaNmyY3nzzzSb3lZmZqczMTEnSnDlz5PF4Wqboo1RdXqJCSfGJXdSpldWGtsvlcrW6vo62jT6FYKI/IdjoUwi29tqnQhL2jDENllmW5Zu2bVtLly7VtGnTDruv9PR0paen++YLCgqCU2SQmNp6SkrLdLCV1Ya2y+PxtLq+jraNPoVgoj8h2OhTCLZI7lM9e/Zscl1Iwp7b7VZhYaFvvrCwUImJib758vJyff/993r44YclSfv27dMTTzyh//f//p9SUlJCUWLwcM0eAAAAgFYgJGEvJSVFeXl5ys/PV1JSktasWaPp06f71sfExGjx4sW++YceekjXXntt2wt6EtfsAQAAAGgVQpJInE6npk6dqtmzZ8u2bY0bN059+vRRRkaGUlJSlJaWFooyQoORPQAAAACtQMiGn4YOHaqhQ4f6LZs4cWKjbR966KEQVNRCGNkDAAAA0AqE7KHq7QYjewAAAABaAcJesDGyBwAAAKAVIOwFGyN7AAAAAFoBwl6wMbIHAAAAoBUg7AWZYWQPAAAAQCtA2As2RvYAAAAAtAKEvWBjZA8AAABAK0DYCzZG9gAAAAC0AoS9YGNkDwAAAEArQNgLNkb2AAAAALQChL1gY2QPAAAAQCtA2As2RvYAAAAAtAKEvWCrG9lz8KMFAAAAED4kkmCrqZZcLlmWFe5KAAAAALRjhL1gq6nhFE4AAAAAYUfYC7aaGlmEPQAAAABhRtgLttrTOAEAAAAgnAh7wcbIHgAAAIBWgLAXbNXVkotn7AEAAAAIL4aggu34fuqQ0EWV4a4DAAAAQLsWsrCXlZWlJUuWyLZtjR8/XhMmTPBb/+677+qdd96Rw+FQx44ddcstt6h3796hKi9oHOdcpM4ejwoKCsJdCgAAAIB2LCRhz7ZtLV68WLNmzZLb7dZ9992ntLQ0vzA3atQonXfeeZKk9evXa+nSpZo5c2YoygMAAACAiBOSa/ZycnKUnJys7t27y+VyaeTIkVq3bp1fm5iYGN90eXk5DyUHAAAAgGMQkpG9oqIiud1u37zb7VZ2dnaDdm+//bb+9a9/qbq6Wg888ECj+8rMzFRmZqYkac6cOfJ4PC1T9DFwuVytsi60XfQpBBt9CsFEf0Kw0acQbO21T4Uk7BljGixrbOTu/PPP1/nnn69PPvlEy5Yt0x133NGgTXp6utLT033zrfHaOA/X7CHI6FMINvoUgon+hGCjTyHYIrlP9ezZs8l1ITmN0+12q7Cw0DdfWFioxMTEJts3dponAAAAACBwIQl7KSkpysvLU35+vqqrq7VmzRqlpaX5tcnLy/NNb9iwQT169AhFaQAAAAAQkUJyGqfT6dTUqVM1e/Zs2batcePGqU+fPsrIyFBKSorS0tL09ttv66uvvpLT6VRcXJxuv/32UJQGAAAAABEpZM/ZGzp0qIYOHeq3bOLEib7pG264IVSlAAAAAEDEs0xjd08BAAAAALRpIblmr7259957w10CIgx9CsFGn0Iw0Z8QbPQpBFt77VOEPQAAAACIQIQ9AAAAAIhAzoceeuihcBcRifr27RvuEhBh6FMINvoUgon+hGCjTyHY2mOf4gYtAAAAABCBOI0TAAAAACIQYQ8AAAAAIlDIHqreXmRlZWnJkiWybVvjx4/XhAkTwl0SWrmCggItXLhQ+/btk2VZSk9P1wUXXKCSkhLNmzdPe/fuVdeuXfXrX/9acXFxMsZoyZIl2rhxozp06KBp06a1y3PQcXi2bevee+9VUlKS7r33XuXn5+vpp59WSUmJTjjhBN15551yuVyqqqrSggUL9O233yo+Pl533323unXrFu7y0cocPHhQzz//vL7//ntZlqXbbrtNPXv25HsKR+V///d/tXLlSlmWpT59+mjatGnat28f31EI2B//+Edt2LBBCQkJmjt3riQd1e9Oq1at0j/+8Q9J0mWXXaaxY8eG6yO1CEb2gsi2bS1evFgzZszQvHnztHr1auXm5oa7LLRyTqdT1157rebNm6fZs2frnXfeUW5urt544w2deuqpmj9/vk499VS98cYbkqSNGzdq9+7dmj9/vm6++WYtWrQozJ8ArdVbb72lXr16+eb/+7//WxdeeKHmz5+v2NhYrVy5UpK0cuVKxcbG6tlnn9WFF16oV155JVwloxVbsmSJBg8erKefflp/+MMf1KtXL76ncFSKioq0YsUKzZkzR3PnzpVt21qzZg3fUTgiY8eO1YwZM/yWHel3UklJif7+97/rscce02OPPaa///3vKikpCflnaUmEvSDKyclRcnKyunfvLpfLpZEjR2rdunXhLgutXGJiou+vS506dVKvXr1UVFSkdevWacyYMZKkMWPG+PrS+vXrNXr0aFmWpdTUVB08eFDFxcVhqx+tU2FhoTZs2KDx48dLkowx2rx5s0aMGCHJ+z/J+n2q7i+ZI0aM0Ndffy3u3YX6SktLtXXrVp1zzjmSJJfLpdjYWL6ncNRs21ZlZaVqampUWVmpLl268B2FI9K/f3/FxcX5LTvS76SsrCwNGjRIcXFxiouL06BBg5SVlRXyz9KSOI0ziIqKiuR2u33zbrdb2dnZYawIbU1+fr527Nihfv366ccff1RiYqIkbyDcv3+/JG8/83g8vm3cbreKiop8bQFJeumll3TNNdeorKxMknTgwAHFxMTI6XRKkpKSklRUVCTJ/7vL6XQqJiZGBw4cUOfOncNTPFqd/Px8de7cWX/84x/13XffqW/fvrr++uv5nsJRSUpK0sUXX6zbbrtN0dHROu2009S3b1++o3DMjvQ76dDf3ev3u0jByF4QNfZXJsuywlAJ2qLy8nLNnTtX119/vWJiYppsRz/D4XzxxRdKSEgI+Bop+hQOp6amRjt27NB5552nJ554Qh06dPCdHtUY+hSaU1JSonXr1mnhwoX605/+pPLy8mZHU+hPOFZH0ocirW8xshdEbrdbhYWFvvnCwkL+iomAVFdXa+7cuTr77LN1xhlnSJISEhJUXFysxMREFRcX+/6C6Xa7VVBQ4NuWfoZDffPNN1q/fr02btyoyspKlZWV6aWXXlJpaalqamrkdDpVVFSkpKQkST99d7ndbtXU1Ki0tLTBqTFo39xut9xut0488URJ3lPp3njjDb6ncFS++uordevWzddfzjjjDH3zzTd8R+GYHel3UlJSkrZs2eJbXlRUpP79+4e87pbEyF4QpaSkKC8vT/n5+aqurtaaNWuUlpYW7rLQyhlj9Pzzz6tXr1666KKLfMvT0tL04YcfSpI+/PBDnX766b7lH330kYwx2r59u2JiYvglCn6uvvpqPf/881q4cKHuvvtuDRw4UNOnT9eAAQP02WefSfLefazu+2nYsGFatWqVJOmzzz7TgAEDIu4vmzg2Xbp0kdvt1q5duyR5f1nv3bs331M4Kh6PR9nZ2aqoqJAxxtef+I7CsTrS76TBgwdr06ZNKikpUUlJiTZt2qTBgweH8yMEnWW4wjWoNmzYoKVLl8q2bY0bN06XXXZZuEtCK7dt2zY98MAD+tnPfub7n9ekSZN04oknat68eSooKJDH49FvfvMb3+2DFy9erE2bNik6OlrTpk1TSkpKmD8FWqvNmzfrzTff1L333qs9e/Y0uK15VFSUKisrtWDBAu3YsUNxcXG6++671b1793CXjlZm586dev7551VdXa1u3bpp2rRpMsbwPYWj8vrrr2vNmjVyOp06/vjjdeutt6qoqIjvKATs4/c6TAAACNpJREFU6aef1pYtW3TgwAElJCToyiuv1Omnn37E30krV67U8uXLJXkfvTBu3LhwfqygI+wBAAAAQATiNE4AAAAAiECEPQAAAACIQIQ9AAAAAIhAhD0AAAAAiECEPQAAAACIQIQ9AEDIPPbYY77nZR2trVu36q677gpKPQ899JDef//9oOwLAIDWhkcvAACO2u233659+/bJ4XCoY8eOGjJkiKZOnaqOHTuGu7SAPPTQQzr77LM1fvz4Rtfv2rVLf/3rX7V582ZVV1era9euGjt2rC644AI5HJH/99KFCxfK7XbrqquuCncpAICj4Ap3AQCAtu13v/udBg0apKKiIs2ePVvLli3T5MmT/doYY2SMaVMBaffu3Zo5c6bGjh2rJ598UomJidq1a5f+9re/qaysTLGxseEuEQCAZhH2AABBkZSUpMGDB+v777+X5B01O+n/t3f/MTW+fQDH386pkzipTkdrkWWTqCUZVskUWYZmKqzZOJuKjWwZaav5C41WW2WtUZZamJLMzGyalmgNExGNYX4cQzqVu3LqOH3/aM70kHqY70PP5/XXfZ/rvn7d9c/nfK7rOr6+tLS08PTpU7KzsyksLLRl0mpra6mpqcHHx4erV68yYcIEEhISCAoKAkBRFEpLS7l79y59fX3Mnj2b1NRUHjx4QH5+PoWFhcBgdjEyMpK6ujo6OjpYsGABCQkJaDQaFEXhyJEjPH78GKvViq+vL4mJibi5uY04nzNnzjBz5kw2b95s+8zT03PIEtJbt25x8uRJ2tvb8fb2JiEhgalTp9rGFRUVRV1dHW/fviU0NJT4+HgKCgp49OgRPj4+pKSkoNVqeffuHTt27CApKYmKigoGBgaIjo4mOjoagP7+fsrLy2loaAAgJCSEjRs3Ym9vb3sfq1at4vz586hUKuLj44mIiLDVPXXqFA0NDVgsFhYsWIDBYECj0fyw7pUrV6ivrwfg4sWL+Pv7k5aWRnV1NZcuXaK3txdXV1cSEhIICAj4pf8dIYQQv8ff8xWrEEKIP1pbWxt37tzB29vb9lldXR1JSUmUlpai1+u/qfPkyRM8PT0pLi5mzZo1FBYW8mV3QX5+PmazmezsbI4dO8bq1auH7bu+vp709HTy8/N58+YNVVVVwGBGMTw8nIKCAgoKCtBoNBQXF49qPs3NzQQHBw9bbjQayc3NxWAwUFRURFBQEIcOHcJisdieaWxsJCMjg9zcXG7fvk1mZibx8fEUFxdjtVq5dOnSkDbv379Pbm4uGRkZVFdXc+/ePQCqqqp4/Pgxhw8fJisriydPnnD27FlbvY6ODnp6eigsLGTbtm0UFxejKAoA5eXlvHnzhqysLPLy8mhvb6eysnLEupGRkYSFhbFmzRrKyspIS0vDaDRy+fJlMjMzKS0tJT09ncmTJ4/qfQohhPj3SbAnhBDil2RlZWEwGNi3bx9+fn7ExMTYysLDw/Hy8kKtVmNn9+1iEr1eT2RkJCqViiVLlmAymejs7MRkMtHU1ERiYiJarRY7Ozv8/PyGHUNUVBR6vR6tVsvatWu5fv06AE5OTgQHB+Pg4ICjoyMxMTE8fPhwVPNSFAVXV9dhy2/cuEFQUBBz5szBzs6O6Oho+vr6aG1ttT2zYsUKXFxc0Ol0zJo1ixkzZjB9+nTs7e1ZuHAhz549G9LmunXrGD9+PNOmTSMiIsI2j/r6emJjY3F2dmbSpEnExcVx7do1Wz21Wk1cXBx2dnbMmzeP8ePHYzQaGRgYoKamhs2bN6PVam3v4Eu7P6r7PSqViv7+fl69eoXFYsHd3R0PD49RvU8hhBD/PlnGKYQQ4pfs2bOHOXPmfLdspOWSLi4utmsHBwcAPn36hKIoaLVatFrtqMbwddZw8uTJtLe3A2A2mzlx4gRNTU10d3cD0Nvbi9VqHXH/oFarxWQyDVtuMpmGZLVUKhV6vd7WN4Czs7PtWqPRfHNvNpuHtPn1+9Lr9bx48QKA9vb2IX19PUcYDGrVarXt3sHBgU+fPtHV1YXZbCYtLc1WNjAwgNVqHbHu93h4eGAwGKioqODVq1cEBgayadMmdDrdd58XQgjxvyXBnhBCiN9m3LhxP1XPzc0NRVHo7u4e1UEobW1tQ66/BB8XLlzAaDRy8OBBXFxceP78OampqYzmIOqAgAAaGxtte9/+k6urqy0Yg8Eg6uu+f8aHDx+YMmWKbR5fMos6nY7379/j5eX1zRx/xMnJCY1GQ05Ozk+N63t/v7CwMMLCwujp6eHo0aOUl5eTnJz8X7cthBDi95NlnEIIIf44rq6uzJ07l6KiIhRFwWKx0NLSMuzzly9f5sOHDyiKwrlz5wgJCQEGs4QajYYJEyagKAoVFRWjHsP69etpbW2lrKyMjo4OYPCEzry8PLq7uwkNDeXOnTs0NzdjsVi4cOEC9vb2+Pr6/vS8z549i9ls5uXLl9TW1hIaGgrAokWLqKqqoquri66uLiorK1m8ePGI7alUKpYtW0ZJSQmdnZ3AYJawqalpVONxdnbm7du3tnuj0cj9+/fp7+9Ho9Gg0Wj+qhNWhRDi/41k9oQQQvyRkpOTKSkpISUlBYvFgr+//7D79sLCwti/fz8mk4n58+cTGxsLwMqVK8nLy2PLli3odDpWr17NzZs3R9W/h4cHBw4c4PTp0+zatYvPnz/j7u5OeHg4jo6OTJw4keTkZI4fP247jXPv3r3f3Zs4Wn5+fuzcuROr1Up0dDSBgYEAxMTE0NPTw+7duwEIDg4esjfyRzZu3EhlZSXp6el8/PgRnU7H8uXLmTt37oh1ly5dSk5ODgaDAT8/PzZs2EB5eTmvX79GrVbj6+tLUlLST89XCCHE7yU/qi6EEOKvtn37drZu3TrsvsG/wZefXjh16tSQ/XNCCCHEr5C1F0IIIYQQQggxBkmwJ4QQQgghhBBjkCzjFEIIIYQQQogxSDJ7QgghhBBCCDEGSbAnhBBCCCGEEGOQBHtCCCGEEEIIMQZJsCeEEEIIIYQQY5AEe0IIIYQQQggxBv0DamOYX2b01bsAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1080x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "x_train_processed = (x_train_processed - x_train_processed.mean()) / x_train_processed.std()\n",
    "x_test_processed = (x_test_processed - x_test_processed.mean()) / x_test_processed.std()\n",
    "cov_mat = np.dot(x_train_processed.T, x_train_processed)\n",
    "eigenvalues, eigenvectors = np.linalg.eig(cov_mat)\n",
    "\n",
    "# Each feature explained variance\n",
    "explained_variance = [ev / np.sum(eigenvalues) for ev in eigenvalues]\n",
    "# Cumulative Variance\n",
    "cumulative_variance = [np.sum(explained_variance[:i+1]) for i in range(len(explained_variance))]\n",
    "\n",
    "plt.style.use('ggplot')\n",
    "plt.figure(figsize=(15, 4))\n",
    "plt.plot(cumulative_variance)\n",
    "plt.title('PCA Cumulative Variance')\n",
    "plt.xlabel('Principal Components')\n",
    "plt.ylabel('Cumulative Variance')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(50000, 200)"
      ]
     },
     "execution_count": 115,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train_processed_pca = np.dot(x_train_processed, eigenvectors[:200].T)\n",
    "x_test_processed_pca = np.dot(x_test_processed, eigenvectors[:200].T)\n",
    "x_train_processed_pca.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/500\n",
      "313/313 [==============================] - 5s 11ms/step - loss: 1.9861 - accuracy: 0.2875 - val_loss: 1.8331 - val_accuracy: 0.3552\n",
      "Epoch 2/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 1.8122 - accuracy: 0.3579 - val_loss: 1.7617 - val_accuracy: 0.3795\n",
      "Epoch 3/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 1.7155 - accuracy: 0.3986 - val_loss: 1.6780 - val_accuracy: 0.4171\n",
      "Epoch 4/500\n",
      "313/313 [==============================] - 3s 11ms/step - loss: 1.6297 - accuracy: 0.4263 - val_loss: 1.6370 - val_accuracy: 0.4231\n",
      "Epoch 5/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 1.5553 - accuracy: 0.4507 - val_loss: 1.5981 - val_accuracy: 0.4395\n",
      "Epoch 6/500\n",
      "313/313 [==============================] - 3s 11ms/step - loss: 1.4911 - accuracy: 0.4775 - val_loss: 1.5665 - val_accuracy: 0.4433\n",
      "Epoch 7/500\n",
      "313/313 [==============================] - 2s 6ms/step - loss: 1.4316 - accuracy: 0.4979 - val_loss: 1.5656 - val_accuracy: 0.4443\n",
      "Epoch 8/500\n",
      "313/313 [==============================] - 2s 6ms/step - loss: 1.3751 - accuracy: 0.5163 - val_loss: 1.5541 - val_accuracy: 0.4555\n",
      "Epoch 9/500\n",
      "313/313 [==============================] - 2s 6ms/step - loss: 1.3147 - accuracy: 0.5384 - val_loss: 1.5585 - val_accuracy: 0.4555\n",
      "Epoch 10/500\n",
      "313/313 [==============================] - 3s 8ms/step - loss: 1.2681 - accuracy: 0.5536 - val_loss: 1.5509 - val_accuracy: 0.4635\n",
      "Epoch 11/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 1.2064 - accuracy: 0.5749 - val_loss: 1.5813 - val_accuracy: 0.4502\n",
      "Epoch 12/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 1.1604 - accuracy: 0.5917 - val_loss: 1.5656 - val_accuracy: 0.4639\n",
      "Epoch 13/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 1.1106 - accuracy: 0.6077 - val_loss: 1.5580 - val_accuracy: 0.4706\n",
      "Epoch 14/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 1.0558 - accuracy: 0.6309 - val_loss: 1.5752 - val_accuracy: 0.4707\n",
      "Epoch 15/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 1.0082 - accuracy: 0.6464 - val_loss: 1.6092 - val_accuracy: 0.4670\n",
      "Epoch 16/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 0.9596 - accuracy: 0.6618 - val_loss: 1.6136 - val_accuracy: 0.4674\n",
      "Epoch 17/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 0.9173 - accuracy: 0.6758 - val_loss: 1.6424 - val_accuracy: 0.4608\n",
      "Epoch 18/500\n",
      "313/313 [==============================] - 3s 11ms/step - loss: 0.8713 - accuracy: 0.6941 - val_loss: 1.6663 - val_accuracy: 0.4668\n",
      "Epoch 19/500\n",
      "313/313 [==============================] - 3s 11ms/step - loss: 0.8269 - accuracy: 0.7071 - val_loss: 1.7204 - val_accuracy: 0.4608\n",
      "Epoch 20/500\n",
      "310/313 [============================>.] - ETA: 0s - loss: 0.7825 - accuracy: 0.7236Restoring model weights from the end of the best epoch: 10.\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 0.7837 - accuracy: 0.7232 - val_loss: 1.7201 - val_accuracy: 0.4657\n",
      "Epoch 20: early stopping\n",
      "######################\n",
      "Training Time: 61.53s\n",
      "batch size: 128\n",
      "n_layers: 4\n",
      "optimizer: <keras.optimizer_v2.adam.Adam object at 0x00000175BA668358>\n",
      "activation_function: relu\n",
      "loss_function kl_divergence\n",
      "Model: \"sequential_2\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " dense_10 (Dense)            (None, 1024)              205824    \n",
      "                                                                 \n",
      " dropout_8 (Dropout)         (None, 1024)              0         \n",
      "                                                                 \n",
      " dense_11 (Dense)            (None, 512)               524800    \n",
      "                                                                 \n",
      " dropout_9 (Dropout)         (None, 512)               0         \n",
      "                                                                 \n",
      " dense_12 (Dense)            (None, 256)               131328    \n",
      "                                                                 \n",
      " dropout_10 (Dropout)        (None, 256)               0         \n",
      "                                                                 \n",
      " dense_13 (Dense)            (None, 128)               32896     \n",
      "                                                                 \n",
      " dropout_11 (Dropout)        (None, 128)               0         \n",
      "                                                                 \n",
      " dense_14 (Dense)            (None, 10)                1290      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 896,138\n",
      "Trainable params: 896,138\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "313/313 [==============================] - 1s 3ms/step - loss: 1.5508 - accuracy: 0.4600\n",
      "Test Loss: 1.5507643222808838 Test Accuracy: 0.46000000834465027\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.51      0.43      0.47      1000\n",
      "           1       0.54      0.56      0.55      1000\n",
      "           2       0.37      0.37      0.37      1000\n",
      "           3       0.31      0.32      0.31      1000\n",
      "           4       0.40      0.35      0.37      1000\n",
      "           5       0.44      0.35      0.39      1000\n",
      "           6       0.44      0.58      0.50      1000\n",
      "           7       0.52      0.54      0.53      1000\n",
      "           8       0.54      0.62      0.57      1000\n",
      "           9       0.52      0.47      0.50      1000\n",
      "\n",
      "    accuracy                           0.46     10000\n",
      "   macro avg       0.46      0.46      0.46     10000\n",
      "weighted avg       0.46      0.46      0.46     10000\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAaQAAAFRCAYAAAA7E9mfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOzdd3hM2f/A8fdMJj1CEiR6772sLkpC9N5WWYRltegEq+8SYpXohFgi2iK62OhtCYkSBCFKSEiVIn3u74/5GqIlYkbC77yeZ54nueV8zk3mzueec8+cK5MkSUIQBEEQspk8uysgCIIgCCASkiAIgpBDiIQkCIIg5AgiIQmCIAg5gkhIgiAIQo4gEpIgCIKQI4iEJORIqamp2NvbY2FhgUwm49SpUxopt3jx4vzxxx8aKSunGzBgALa2ttldDUHINJGQhEyLiIhg0qRJlCtXDgMDA/Lnz4+1tTWbN28mNTVVo7F2796Nh4cHBw4cICQkhAYNGmikXB8fH8aOHauRsj7n1KlTyGQy9PX1CQ8PT7cuJSUFS0tLZDIZ7u7umS7z3LlzyGQyHj16lKntly1bxq5du76k2oKQrRTZXQHh+xAcHEzDhg1RKBTMmTOHGjVqoKury4ULF1i0aBFVq1alevXqGot3//59ChUqpLFE9Ea+fPk0Wl5GrKys2Lx5M+PGjVMv27t3L4aGhlqLmZKSgkKhIHfu3FqLIQhaIQlCJrRr106ytLSUoqOjP1iXnJwsxcXFqX+ePHmyVLBgQUlXV1eqUKGCtHXr1nTbA9LKlSulvn37SiYmJlLhwoWlBQsWqNc3adJEAtSvYsWKqZcPGjQoXVlz585Vr5ckSfL395datmwp5c6dWzIyMpLKly8vbd68Wb2+WLFi0ty5c9W/x8TESEOGDJHy5s0r6evrS7Vq1ZK8vLzU64OCgiRA2rFjh9SuXTvJ0NBQKlGiRLoyP+bkyZMSIM2ePVsqX758unU2NjbSnDlzJEDasmWLevnSpUulatWqScbGxpKlpaXUs2dP6fnz5+nq8e6rSZMmkiRJUv/+/SUbGxvJxcVFKlasmCSTyaTY2Fj1ckmSpMTERKl69epSx44d1fFev34tVapUSerRo8dnj0UQvhXRZSdkKDIyksOHDzNy5MiPXnXr6upibGwMwNSpU1m/fj1Lly7F39+fvn370rdvX44fP55un9mzZ2Ntbc21a9eYOHEikydP5uTJkwDs2bOH8ePHU7x4cUJCQvDx8cl0XX/++WcsLCy4cOECN2/eZPHixZiZmX1ye3t7e7y8vHB3d8fPz4+GDRvSrl07AgIC0m3n6OhIv379uHHjBj169GDgwIHcv38/w/r06tWLkJAQzp07B8CDBw84ffo09vb2H91+0aJF3Lx5k7179/LkyRN69eoFQJEiRdi3bx8Aly9fJiQkhD179qj3u3z5MidOnMDT05Pr169jYGCQrlx9fX127NjB8ePHWbFiBQAODg68fv2adevWZXgcgvBNZHdGFHK+S5cuSYC0e/fuz24XHx8v6enpSStXrky3vFOnTlKzZs3UvwPSqFGj0m1Trlw5ydHRUf37zJkzpVKlSqXbJjMtJFNTU8nNze2TdXy3hXT//n0JkA4dOpRumxo1akgDBw6UJOlty+Svv/5Sr09JSZGMjY2lNWvWfDLOmxbS06dPpWHDhkm//PKLJEmSNHnyZKl9+/bqv8O7LaT3+fr6SoAUHBwsSZIknT17VgKkoKCgdNv1799fyp07txQbG/vB8jctpDc2bdok6evrS9OnT5d0dXWlS5cufTK+IHxrooUkZEj63/y7Mpnss9sFBgaSnJyMtbV1uuVNmjTh1q1b6Za9f7+pUKFCvHjx4qvrOmHCBAYPHkzTpk2ZNWsWvr6+n9z29u3bAB/U19ra+rP1VSgUWFpaZrq+Q4cOZdeuXYSFhbFp0yZ+/fXXj2536tQp7OzsKFKkCLly5aJRo0YAPH78OMMYFSpUwMTEJMPt+vfvT8eOHZk7dy5z586lTp06mToGQfgWREISMlSmTBnkcvkHH9Kf8n7ikiTpg2V6enof7KNUKj9brlwuVyfHN1JSUtL9Pn36dO7du0ePHj3w9/enXr16/P7775mqt6br+0a1atWoXLkyP//8MwqFgjZt2nywzZMnT2jTpg3Fixdn+/btXLlyhf379wOQnJycYYw3XaYZiYuLw9fXFx0dHe7du5epfQThWxEJSciQubk5rVu3ZsWKFbx69eqD9SkpKcTHx1O6dGn09fU5ffp0uvVnzpyhUqVKX12P/Pnz8/z583TLPtYCKlmyJMOHD+eff/5hzpw5rF69+qPlvanTmTNn0i0/e/asRur7rqFDh3L8+HHs7e3R0dH5YL2Pjw8JCQksXbqUhg0bUq5cuQ9aYG+SYlpaWpbrMWzYMHR0dDhx4gTu7u5s3749y2UJgqaJhCRkyqpVq9DV1aVWrVp4eHhw+/ZtAgMDcXd3p3bt2ty/fx8jIyMcHByYPn06u3bt4v79+8ybN499+/YxderUr66Dra0t3t7e7Ny5k8DAQJycnDh79qx6fVxcHCNGjODEiRMEBQXh5+fH0aNHqVix4kfLK1WqFN27d2f48OF4eXkREBDA6NGj8ff3Z+LEiV9d33cNGDCAsLAwpk+f/tH1ZcqUQSaT8ddffxEUFISnpydz5sxJt02xYsWQy+UcPnyYly9ffvTi4HPc3d3ZtWsX27dvx9ramnnz5jF06FCCgoKyfFyCoEkiIQmZUrRoUXx9fenYsSOzZs2iZs2aNGjQgPXr1zNx4kQqV64MwJ9//smvv/7KmDFjqFSpEu7u7ri7u2NjY/PVdejfvz8jRoxg5MiR1K5dm6dPn+Lg4KBer1AoiIqKYtCgQVSoUAE7OzssLS3x8PD4ZJmurq7Y2dnRt29fqlWrxvnz5zl48CDly5f/6vq+S0dHh7x586Krq/vR9VWrVmX58uWsXbuWihUrsmjRIpYuXZpuG0tLS+bPn4+TkxMFChSgY8eOmY4fGBjI8OHDcXZ2Vt8PGzduHA0aNKBXr14fdH0KQnaQSe93yguCIAhCNhAtJEEQBCFHEAlJEARByBFEQhIEQRByBJGQBEEQhBxBJCRBEAQhR/guHj8R69Au22KX2JR932avk6d0tsU+HX4722IDlMhllW2xA6KeZlvsfEbZ98iIxNSMZ4TQpj55a2Vb7B1R17ItNkDYq7saKysl/GGW9tPNW1Jjdcgq0UISBEEQcoTvooUkCIIgZJIy61NLZTeRkARBEH4kUuYm/c2JREISBEH4kWRyFvqsiI+PZ82aNTx9+hSZTMawYcMoWLAgS5YsISwsjHz58jF27FhMTEyQJAk3Nzf8/PzQ19dn+PDhlCz5+ftU4h6SIAjCD0SSlFl6ZYabmxvVq1dn6dKlODs7U6hQITw9PalSpQouLi5UqVIFT09PAPz8/AgNDcXFxYUhQ4bg6uqaYfkiIQmCIPxIlMqsvTLw+vVr7ty5Q/PmzQHVZMbGxsb4+PjQpEkTQPUwTh8fHwCuXLmCtbU1MpmMsmXLEh8fT1RU1GdjiC47QRCEH4mW7iG9fPkSU1NTVq1axePHjylZsiQDBgzg1atXmJmZAWBmZkZMTAwAkZGR5M2bV72/hYUFkZGR6m0/RiQkQRCEH0kWR9l5e3vj7e2t/t3W1hZbW1v172lpaQQFBWFvb0+ZMmVwc3NTd899zMceJPH+k5jf9+MkJJkco4lLkKIjSFg3B/2fHdApWgYAZdhzEt2XQHIius06oVu/JaSlIcXFkOixFCkq7KvD6+vrceCoB3p6eigUOhzY58WCeS4cOOqBiYnq8dL58pnje/Umv/Qe/tXxAMY4j6GOTR2iI6IZ3kJVpv1Ue+ra1iU1JZWQxyEsmbCE+Jh4AIqXL86o+aMwymWEpJQY3X40KUlf/xycQoUKsN51MZaW+VAqlbht3MaqVW5MnzGOdm1boJQkwl6GM2ToBEJDXn51vHcVL1WURev+UP9euFghVixch8+5q0x3noyRsSHPn4YyedgM4uNeazT2x4x2+BV7+5+RJAl//wAGDR5HUlKS1uLJ5XKOntpF6PMX/NJrOI2s6zFj7gRkcjmv4+IZPXwaj4KeaDyuvr4eh7y2oa+vh45CwX7Pozj9uYx1G/6ieo0qpKamcvXKdcY6TCc1NfWr4/VZ+BuVm9ckNiKGeXYTABi4YjSWJQsCYGhqRELMa5zaTAagYPmi/DzvVwxMDJGUEgs7TiVVA+91fX099h/Zmu48Xzh/OY2s6zH7j0no6upy49otRo+c9lVP9v0qWWwhvZ+A3mdhYYGFhQVlyqg+V+vVq4enpye5c+cmKioKMzMzoqKiMDU1VW8fHh6u3j8iIuKzrSP4Tp6HlJmZGnSbdUKnSGlkBkYkrJsDBoaQmACAfufBSLHRJHv/g06ZKqQ9ugcpSeg2ao1O6Sokblr4yXK/ZKYGY2Mj4uNfo1AoOHRsG1Mn/8FVn+vq9W5blnPk8HF2bvv0VcW7MpqpoXKdyiS8TmD8kvHqhFSjcQ2uX7iOMk3JwCkDVXHnuyHXkbP88HIWjVlE0J0gcuXJRXxMPMpP9B1/yUwNVlb5sLLKz7VrtzAxMebc+QP06jmEZ89CiY2NA2DYsAGUr1CG0Q7TMlVmVmZqkMvlnLh+gJ9bD2KJ6zwWzV7OlYt+dP65HYWKFmTFgnWZKierMzUULGjF6ZN7qVKtGYmJiWzzWMORIyfYvGVnpsv40pkaho7oT9XqlciVy4Rfeg3n3JXDDOw9kvv3HtJ/UC9q1KrCmOGZ+5t/6UwN777fj/y7nSmT/sDMLDf/HlM9wt7VbQkXzvuw0fXTD0h81+dmaihVpwJJ8Yn8sniEOiG9q/O0fiTEvuaoy27kOnImH3Ri87iVPLvzGOM8JryOiUdSfvqj7ktmanj3uA96eTB9ynzWuy2hS4cBPHzwiMlTHQh++pytW/7JdJmanKkh+eHlLO2nV7JOhtvMmDGD3377jYIFC7Jz5071xVauXLno1KkTnp6exMXF0bdvX3x9fTl69ChTpkzh/v37uLm5MX/+/M+W/00GNTx79gxPT082btyobuYFBwdrrHxZHgsUFX8i5eKxtwv/l4wA0NUDVG/GtPs3IUX1R0x7dBd5nrxoSny86gpcV1eBrkKRrslqYmJMY+t6HD74r8bi+V/2JzY6Nt0yv7N+KNNUSSbAN4C8Vqrjq2ldk6A7QQTdUT2uOjY69pPJ6EuFhoZx7dotAOLi4rl79wEFC1qpkxGoTmJtX/vUa1ybp4+eERIcSvHSxbhy0Q+Ai6cv06JtM63GfkOhUGBoaICOjg5GhoaEhIRqLVaBgpbYtGyCx5bd6mWSJGGSywQAU9NcvAj5+tb/p6R7v+vqIkmSOhkBXL1yg4KFNDMF1IPLd3j9Ku6T62u2rcfV/ecBKN+4Ks8CnvDszmNVPaPjPpuMvlT641aQlpZGcnIyDx88AuD0yfO069BSY/G+lDZH2dnb2+Pi4sKECRN49OgRnTt3plOnTty4cQMHBwdu3LhBp06dAKhRowb58+fHwcGBtWvXMnjw4AzL13qXnaenJ+fPn6dhw4aULq264o+MjGTZsmU0bNhQXfmvod9lCEn7NyLTN0q33KD3aHQq1kb54ikJezd8sJ9uvZak3r761fHfkMvlHD+zlxIli7Jx/VZ8r9xQr2vTvgVnTl8kLjZeY/Ey0rJnS84cOANAoZKFAJi7ZS65zXNz5sAZ/lmT+Su4zCpatDDVqlXEx0d1xTlz1gR69+5CzKtYWrf+WePx3tW6cwsO71VdlAQGPKBZq8acPHqWlu1tsCqUX6uxAZ4/D2XxkjUEPbhMQkIi/3qf5l/vM1qLN2e+I3/MWIRxLmP1sgkOM3DftYbEhETiYuNp26KX1uLL5XJOnfOkRMlibFjnztUrb3sDFAoFPX/uxJRJc7UW/41SdSoQG/6KsEeq5J+/ZEGQJEZsnoqJuSlXD1zAe+1+jcWTy+UcP72HEiWLssHVA9+rN1AoFFSrUZnrfv6079hKY4k4S7T4PaTixYvj5OT0wfIZM2Z8sEwmk2UqCb1L6y2kkydPMn/+fDp16oS1tTXW1tZ06tSJ+fPnc+LEia8uX6fST0ix0SifPvhgXaLHMuKn90cZ+hRFzcbp1ilqN0WnaGmST+z+YL+sUiqVNGvUkaoVrKlZqyrlK5RRr+vSrR17/jmosVgZ6TmyJ2mpaZzcexIAHR0dKtauiLODMxO7TqS+XX2qNaym0ZjGxkZ4bFvNpElz1K2j2bMWUa5sA3bs2MfQ3/prNN67FLoKmrZszLEDqvfU9DF/8vPAbuw4tgljEyNSkr/+PkZG8uTJTYf2dpQuW48ixWpibGxE795dtBLL1q4J4WGR3Lievmt1yPBf6Nv9N2pVas72rXuZ9edkrcQH1fvdukEHKpVrRM3a1ahQ8e37fdGS2Vw4f5mLF65oLf4btTs04Mr+C+rfdXTklPypPJtGL2dxtxlUs/uJsg0qayyeUqmkWeNOVK3YhJo1Vef5EPtx/DFvCl4ndhEXF599949AdQ8pK68cQOsJSSaTfXTseVRU1GdHXHh7e+Po6Iijo+Nny9cpWRFFlboYz9yAwYBJ6JStikG/8W83kJSk+p1Ft1qDt/uUrYZey54krJsLGrjh+r6YV7GcP3cZG1tVEjQzz0PNWlX41+uUxmN9jE03G+rY1MHZwVm9LDwknJuXbhITFUNSYhJXTl6hdGXNzSauUCjw8FjDju2e7N/n9cH6HTv20aljK43Fe19jm/rcuXmXiLBIAIICHzOk52h6thzA4b3HePpYc13En2Jj05igR08ID48kNTWVvZ5HqF+vtlZi1albk5atm3H5xr+s2fAXjazrsmXHaipWLoffVVXLfP/eI/xUp4ZW4r8r5lUs585ewsbWGoBJU0aRN6850xznaT22XEdONbs6+B58m5CiQyMJvHSb+KhYUhKTuXXSjyKVS2g8tuo8v0Rz28Zc8blG+9Z9sGvenYsXfHj44LHG42WaMi1rrxxA6112AwYMYM6cORQoUAALCwsAwsPDCQ0NZdCgQZ/c790RH58b1JB84G+SD/wNgE7pKug170zilr+Q5S2AFB4CgKJSHZQvVB9I8sIlMeg1kterZyLFvdLIMQJYWJiRkppKzKtYDAz0sW7agOVLVTfRO3ZqxbGjp0hK0v70/rWa1KL7sO5M6j6JpMS3o7t8z/jS7bdu6Bvok5KSQuV6lfF0zdzgisxYvXoBd+8Gsnz5267RUqWK8+B//ept29py996HrVhNadO5pbq7DsA8rxmR4aqLnqFjB7Lz771ai/3G0yfPqFu3JoaGBiQkJNK8WSOuXr2e8Y5ZMG/OEubNWQJA/UY/MWzkQAb2GcWNe2coWaoYDx88xrpZfe5r6W9ukdeclJQU9fu9abMGLFu8jn79e2Bj05iO7fpp/Z4hQLlGVXjx8DnRoZHqZbdPX8d2aAd0DfRIS0mldN2KnNxwSCPx3j/PmzRtgMvS9eTNa054eCR6erqMGvMrSxat0Ui8LMkhrZ2s0HpCql69OsuWLSMwMJDISNWbxtzcnNKlSyOXa6mBJpNh2HcsGBgBMpTPg0jcuRIA/Y72oGeA4UBVy0uKCiNh/df3c1ta5WfFmgXo6MiRy+Xs23uEY0dPAdC5a1uWLcncCK8vMWn5JKrWr4qpmSmbL23GfbE7PUb0QFdPlz+3/gnAXb+7rJi6grhXcex13cvSg0uRJIkrJ6/gc8JHI/WoX782vft0xf/mHS7+dxiAWTMX8kv/npQtUxKlUsmTp89wyOQIuy9lYKhPfes6zJ7wtm+7TecW9BrYDQDvw6fYu0373aWXffzYs+cQPpe9SE1N5dq1W6x33ar1uG+kpaUxYfQMXDcvQykpeRUdw9gRv2sllpVlPlatc1a/3/fuOYzX0ZOERQfw9Mlzjp3YBcCB/cdwdlrx1fEGuDhQpl5FTMxyMffiKg4v2cXFnSep1b6BejDDGwkx8ZxwPcik/fOQJLh10o9bJ/2+ug7w5jx3Qi7XQS6XsW/vUf71OsXMuZNoadcUuVzOpg3bOHfmP43EyxIt3kPSth9m2Le2iAf0ZQ/xgL5vTzygL/tocth3kn/WRvLqV26hsTpklZjLThAEQcgRfpyZGgRBEITvustOJCRBEIQfiCTljBFzWSESkiAIwo9EjLITBEEQcgTRZScIgiDkCKKFJAiCIOQIOWTWhawQCUkQBOFHIlpIgiAIQo4g7iFpVwHXW9kWO/zEh1Otfysm1uOyLbbpe4/y+NbuRWt/MtScKLeeccYbaUl2z9RwKTkk22K3yFMx22JrnGghCYIgCDmCaCEJgiAIOYJISIIgCEJOIGZqEARBEHIG0UISBEEQcgQxqEEQBEHIEUQLSRAEQcgRvuMWknhAnyAIgpAj/FAtpEKFCrDe9S8sLfOhVCpx27iNVas2MXXaaAYM7EV4eCQAs2Y6c8zrlEZith6/GCMDPXTkcnTkcrbN/g0Aj3//Y7v3JXTkcqyrl2VsTzsANhw4w94zvsjlMib3bUPDKmU0Uo/3jXb4FXv7n5EkCX//AAYNHkdSUpJWYunr63HwqAf6+nooFAr2ex7FaZ4LANNmjKVj59Yo05RsdPVg3ZrNGo+/bu0i2rSxJSwsnBo1bdOtGzt2KAucplOgYBUiIqI0Hvt9o0YOYtCg3shkMjZs8MBluavWYpUoVYzF6+epfy9SrCAuC9ZhWSAfzVo2JiUlhSePgpnqMIfYmDiNxtbX1+OQ1zb09fXQefM//3MZLivnU6NmZWQyGYGBjxgxdBLx8a81Ghug16BudOrTDplMhufWg2xz3cVvEwdhbdcISVISGR7N7DHzCH8RoZF4Q5xHUqN5bWIiXjG55WgAuo7pSbOfWxATEQPATmd3rp30BaDD8C407WmLMk3J5lmu3DjzDR+RLrrscobUtFSmTPmT69duYWJizNnzBzhx4hwAK5ZvxGXZeq3EdXUciFmut9+wv3znIad8A/jnjxHo6SqI+N+HwYNnLzl66SZ75o3kZXQsQxdsYv/C0ejINdtQLVjQipEj7KlSrRmJiYls81hDzx4d2bxlp0bjvJGUlEyndr8QH/8ahULBkWPb8f73DGXLlaJQoQLUrWmHJEnkzWuulfibt+xi1epNuG1cmm554cIFsLFpzOPH32bWh0qVyjFoUG/qN2hLcnIKhw9u5fCR4wQGBmklXtCDx3Ru3gcAuVzO6RuH8T58khKli7H4j5WkpaUxfvpIhowewF9zV2g0dlJSMh3b9nv7P/93O97HTjPN8U9iY1Xv9z/mT+XXof1YunitRmOXKleCTn3a0b/tUFKTU3HxcObc8YtsWb2NNc4bAOg5qCuDxw7AyfEvjcQ8s+sEx/4+zLDFo9MtP7LhAIfW7Uu3rFCZwtRv34hJLRwwszRn6tbZjGs6AulbJQrRZZczvAgN4/o11TRDcXHx3L0bSIGCVt+8HruO+2DfrjF6uqp8b2FqAsAp3wBa1a2Cnq6CwvnMKGJpjv9D7XxYKhQKDA0N0NHRwcjQkJCQUK3EeePNVbCurgKFrgJJkhg46GecF6xAkiQAdQtV086du0RUVPQHyxc5z2LqlD/V8bWtfPkyXLrkS0JCImlpaZw5+x+dOrb6JrHrW//E00fBPA8O5fypS6Slqb6Lcv2qP1YFLbUS893/ua6uLpIkqZMRgKGhvlb+9sXLFOOm722SEpJIS0vD9+I1mrZuTHzc25aYoaGBRmMHXL5NXHRsprat1aIOFw+cIzU5lbCnL3nxKITS1bXTE/JRSmXWXjlAtiakkydPaq3sokULUa1aRa74qJrKQ3/7hf8uHWHVmgXkyWOq0Vi/OW+m14zV/HPyCgCPX0Tge/cxfWavxX7eBvwfPgPgRVQMlua51ftZmufmZVTm3uRf4vnzUBYvWUPQg8sEP/HjVUwM/3qf0Xicd8nlck6f38/dh/9x6uR5rl65TomSRencpS3HT+9h525XSpYqptU6vKtduxY8ex7KjZt3vlnMW7cCaNy4HubmZhgaGtC6VXMKFy74TWK36dSSQ3u8Plje9ecOnDl+QSsx5XI5Zy7s517QJU6dOMfVK9cBWLHaibsP/6NM2VJa6aJ9EBBEjbrVyG1mir6hPg2a18OyYH4Ahk0ezMEr/9CqSwvW/q+1pE0tf2mD09ElDHEeibGpqpfE3MqCiJC3XYURoRGYWWmnd+CjRELKmp07tdOFZGxsxNZtq5k8aS6xsXG4rt9KlUpNqF+vDS9Cw5jnNE1jsf7+fTA75gxj5YR+7Dh+iasBj0hNUxLzOgH3GUMY29OOiSt3fPJqTaaxmryVJ09uOrS3o3TZehQpVhNjYyN69+6ihUhvKZVKmjTsQOXyjalZqyoVKpRBT0+PpKQkbJp0YfPfO1m+ar5W6/CGoaEBjpMdmD170TeJ90ZAQCDOzis5emQbhw9u5fqN26Slav9b87q6CprbWXP0wPF0y4eOGUhqWioH/jmilbhKpRLrBh2oVK4RNWtXo0JFVStg5DBHKpRuwL27D+jcta3G4z4KfMzmVR6s2L4Yl62LuH/7gfrvvHqBK+1qd+Ponn/pYa/d9/y/7kcZYz2MKa3HEf0yij7TB6pWyD48q79VK10VTJm1Vw6g9XtIEyZM+OhySZJ49erVJ/fz9vbG29v7i+MpFAq2eqxmx/Z97N+numJ8+TJcvd5t4zb+2a25K6f8ZqrWloWpCc1rVcD/YTCW5qbY1KqITCajSqnCyGUyomJfY2lmyovIt8f8IvIV+cxyaawub9jYNCbo0RN1F9lezyPUr1cbD489Go/1vphXsZw/ewmbFtY8fx6q/h8c3H+MFau+zczppUoWp3jxIlzxOQao7iVd+u8oDRu148WLMK3Gdtu0HbdN2wH4Y64jwcHan8G6sU0Dbt8MICLsbZdop55tadayEQO6Dtd6/JhXsZw7ewkbW2vu3L4PqJLVnt2HcBg9GA/33RqPuX/bIfZvOwTAcMdfeRmS/v96dK83S7csYN0iN43HfiMm/O25fGLbMSZu/B2AyJBwLApYqNdZWFkQ/UL7A2rUckhrJyu03gW9HtoAACAASURBVEJ69eoVI0eOZPLkyR+8cuX69Iexra0tTk5OODl92YfYqtULuHs3kBXL3yYdS6t86p/bd7Dj9u17X34gH/E6KZn4hCT1zxf9H1C6sCXNalbg8p2HADwKDSclLQ2zXEY0qVGeo5dukpySSnBYFE9eRFK5ZGGN1OVdT588o27dmhgaGgDQvFkjAgLuazzOGxZ5zTHNrfpfGhjo06RZA+7de8jhg95YN6kPQMNGdbR2c/99/rcCKFykOmXL1adsufoEB4dQt14rrScjgHz5VB9ERYoUpFOn1mzf4an1mG0723FozzH1742a1WfwyF8Y1m88iQnaGVn5/v+8abMGBN4PokTJt92yrVo35969h1qJb2aRBwDLQvlp1sYaL09vipR4ey5Z2zXkUeATrcR+I09+M/XPP9nVI/juYwCu/utD/faNUOgpyFckP1YlChB4TXvn3wdEC+nTatasSWJiIsWLF/9gXcWKmn0GSf36tendpwv+NwO48J/q6mnWTGe6d+9A1aoVkCR4/CQYh1FTNRIv8lUcY122AZCapqRN/ao0rFqGlNRUZrh60mXqCnQVOsz9tQsymYzShfPTsk5lOk9Zjo6OnKn92mp8hB3AZR8/9uw5hM9lL1JTU7l27RbrXbdqPM4blpb5WLV2ITo6cuRyOZ57jnDs6En+u3iFdRsWM2zEAOLjXzN6pOa6St+1ZfMKrK3rkzevOQ8f+DBn7l9s+l8r5VvbtWM95hZmpKSk4uAwjejoT/cCaIKBoT4Nm9Rh5oS3w7+nO01ET0+PjbtWAnD96k1mTdRs69TKMh+r1jmr/+d79xzG6+hJjhzbTi5TE2QyGf437zB+zEyNxn1jgetccpvlJjUllYVTlxD7Ko7fF02mWKkiKJUSoc9CmT9ZMyPsAEa6jKNC/UrkMjNl+X/r2b1kOxXqVaZYxRIgSYQFv2TD1DUAPLv/lP8OXcDZezlpqWm4TV//7UbYwXfdQpJJ37RzM2tMjEpkW2zxgL7sEZeckG2xldl4SpTO820GQXzMi9ffsFvpI0qbZt+xl9G1yHgjLfJ4vFdjZSXsmZfxRh9h2EUzF+pf44f6HpIgCML/e99xC0kkJEEQhB+JSEiCIAhCjpDz78J8kkhIgiAIPxLRQhIEQRByBJGQBEEQhBwhh3ynKCtEQhIEQfiRfMctpB9qtm9BEATh+yVaSIIgCD8SMcpOuyyNzDLeSEvy20zJttjB9b7hM1Te0ykw20IDYK7IvpkiHiZqf867TymhnzfbYqcotT8z+ecExWn3mV2fczP5UbbF1rjvuMvuu0hIgiAIQiaJhCQIgiDkCGKUnSAIgpATSEpxD0kQBEHICUSXnSAIgpAjiC47QRAEIUcQXXaCIAhCjiC67ARBEIQcQSSknCOXqQnzl86gbIVSSBI4OsymqW1DbFs3RalUEhEeyaRRM3kZGq7RuIUKFWDt+kVYWuZDqVSyyW07q1dtonKV8ixd9gfGJsY8eRzMYPuxxMbGaSaonh7my5ch09UFHR2STp0mzm0T5stdkBmpvlgqN8tDyp0Aoqf9jkELW4x7/wyAlJBAzF9LSH3wQCNV6TGoKx37tEMmg31bD7HD9R/KVCrNZKdx6BnokZaahvOUJdy+FvDVscY4j6GOTR2iI6IZ3mI4APZT7alrW5fUlFRCHoewZMIS4mPiyV84P2tPrCX4QTAAd/3usmLqiq+uwxslShVj8fq3j4wuUqwgLgvW8SLkJSMnDqFU2eL0sBuA//U7Gok3dtFY9bEPsx0GwKBpg9Id++Lxi4mPiVfvk69gPtaeWMvWJVvZvXa3RuoB2Xeu6evrceCoB3p6eigUOhzY58WCeS4cOOqBiYkxAPnymeN79Sa/9B6u0dgAa9c607q1DWFhEdSq1QKAqlUrsnz5PAwM9ElNTWP06GlcuXJd47Ez5TueqUEmSTm/9qXy1sz0ts4rZuPznx873T3R1VVgYGiApJSIi1OdoP1/7UXpciWZPiFzz50PS4jO1HaWVvmwssrP9Wu3MDEx5sy5/fzcayhr1y1i2tR5nD93mb6/dKd4scL8MXdJpsq8V6tohtvIDA2REhJARwfzlcuJdVlByu3b6vV55s4m8dx5Er2OoVu5EqmPHiPFxaFXtw4mAwcQ+dvHT9gvmamhZLkSzF09A/u2v5GanMpSj4UsdFzMhHlj2L5uFxdPXqZ+87r0G/4zw7uNyVSZn5upoXKdyiS8TmD8kvHqhFSjcQ2uX7iOMk3JwCkDAXCb70b+wvmZ5TZLvV1mZHWmBrlczukbh+nZaoD6fTd70RQWzlqW6YSU0UwNletWJiE+gQlLJ6gTUk3rmlw7fw1lmhL7KfYAbJy/Ub3PtLXTkJQSd6/d/WxCupfwIlN1fEPT59qr5MxfqBkbGxEf/xqFQsGhY9uYOvkPrvq8TQBuW5Zz5PBxdm7zzFR5ccmJmY7dqFEd4uJes2HDEnVCOnjQHRcXV44dO4WdXTPGj/+Nli17ZrrMxMQnmd42I68X/5ql/YzGrddYHbLqh5pc1cTEmJ/q12Snu+pNmJKSSmxMnPoEATA0MkQbOfhFaBjXr90CIC4unrt3AylY0IrSZUpw/txlAE4eP0eHjq00GldKSFD9oFAgUyjSXR3JDA3Rq1mTpLPnAEjxv4UUpzrpU27dRidfPo3UoXiZotzyvU1SQhJpaWn4XrxGk9aNkSQJ41yqK1YTU2PCXmjmStn/sj+x0bHplvmd9UOZpuqqCPANIK/Vt5+Cp771Tzx9FMzz4FAe3n9E0IPHGo/hf+nDY/c94/v22P0CyFvg7bHXt6tP6JNQHt/TbF2y81wDiI9/DYCurgJdhSJdHBMTYxpb1+PwwX+1EvvcuctERaW/UJUkCVPTXADkzp2LkJAvS+4apZSy9soBvkmX3bNnz4iMjKRMmTIYGBiol1+7do3q1atrLE6R4oWIjIhi4fJZlK9UFv8bd5g71ZmE14mMnzqCzj3bEhsTR59OQzQW82OKFi1E1WqVuOJzjTu379GmrS2HD3nTqUsbChUuoNlgcjkW69ehU6gQrz33knLn7ZW4vnVjkq/6Ir1+/cFuhu3aknTpskaq8DAgiN8mD8bUzJSkxCQaNK9HwI27LJ2xgqXbnBk1YxgymYwhHUZqJF5GWvZsyZkDZ9S/WxWxYvnh5byOe83mRZu5dfmWVuK26dSSQ3u8tFJ2ZrXs0ZLTB04DoG+oT/dh3Znaeypdh3bVaJzsPtfkcjnHz+ylRMmibFy/Fd8rN9Tr2rRvwZnTF4mLjf9MCZo1YcJsDh7cgpPTNGQyOc2adf5msT/wHQ/71noL6fDhwyxcuJAjR44wfvx4fHx81Ou2bdum0VgKhQ6VqpZnq9s/dGjem4T4BH5zUHXf/DVvJY2qtWHfP0foN7iXRuO+y9jYiC0eq3CcNJfY2DiGD5vMkKH9OH1uH7lMjElJTtFsQKWSiEGDCevWHd3yFVCUKKFeZWhjQ8Lx4x/solejOkZt2xC7Zq1GqvAo8AlbVm1j+fZFLN26kPu3H5CamkaX/h1ZNnMlHWv3YNmslUxbPEkj8T6n58iepKWmcXLvSQAiX0bSv15/RrUZxfq565nkMglDE0ONx9XVVdDczpqjBz78e38rvUb1Ii3t7bH3G9+Pva57SXyd+e6ozMruc02pVNKsUUeqVrCmZq2qlK/wdiLiLt3aseefg1qJ+ylDhvRj4sQ5lC5dj0mT5rBmjfM3jZ/Od9xC0npCOn78OAsWLGDSpEnMnDmT3bt3c/jwYYDPNue9vb1xdHTE0dEx07FCnr8k9PlLrvv6A3DkwHEqVSufbpv9u4/Sql3zLBxJxhQKBe4eq9i5Yz8H9quulO/fe0inDv1p0qgj/+w6QFCQ5vqK3yXFxZF87Rp6desAIDM1RbdCeZIu/pe+jiVLYjppIlFTpiHFxGgs/oFth+lvN4RhXUYTEx1DcFAwbbrbcfKwqqVy/MApKlYvn0EpX8emmw11bOrg7PD2wyA1OVXdxRV4M5CQxyEULllY47Eb2zTg9s0AIsIiNV52Zth2s6WOTR0WjlqoXlauRjkGTR3Epgub6DSoEz1H9qR9//YaiZfd59obMa9iOX/uMja2jQEwM89DzVpV+NfrlFbjvq9v3654eh4BYPfug9SuXe2bxn+XpFRm6ZUTaD0hKZVKdTdd/vz5mTVrFn5+fvz999+fTUi2trY4OTnh5OSU6VjhLyMIefaCEqWLAdDAug6Bd4MoXrLI23JbWfPg/qOsHUwGVq524u7dB6xcvkG9LG8+CwBkMhkTJ49gwwYPjcWT5c6NzMRE9YueHvq1apH2WJXwDJo2JeniRUhOVm8vz5+fPH/M5dWf80gLDtZYPQDMLPIAYFkoP03bWHPM8zjhLyKoWV/VJVu7UU2eBmk25rtqNalF92HdmT1oNkmJSerlpuamyOWqt7lVUSsKlihIyOMQjcdv29mOQ3uOabzczKjV9H/Hbp/+2Cd2nciABgMY0GAAnhs82bFiBwf+PqCRmNl5rllYmGGaW3W/xsBAH+umDbh//yEAHTu14tjRUyQlJX+uCI0LCXmBtXU9AJo1a0hg4KNvGv9HofV7SHny5OHRo0cUL14cAAMDAxwdHVm9ejVPnmi+tTB7ygKWrPkTXV1dnj4OZtKoWcxfOoOSpYuhVEo8Cw5h+vg/NR63Xv3a/Ny7C/7+AZy7qOoumDNrEaVKFefXIf0A2L/fC/fNuzQWU8fCgtxTp4COHGRyEk+eVCUhwNCmOXFb0yc/kwH9kec2xXTsWNWCtDQihgzVSF3mu84ht5kpqSmpLJq6lNhXccyfuIixc0aio6NDclIy8yf+pZFYk5ZPomr9qpiambL50mbcF7vTY0QPdPV0+XOr6n/7Znh3lbpV6Du+L2mpaSjTlKyYuoK4Vxoadv8/Bob6NGxSh5nvjCazbdOU3+dNwNzCjDUeSwjwv8fgng5fHWvyislUrVcVU3NTtlzewpa/ttBzZE/VsXuojj3AN0CjQ9s/JbvONUur/KxYswAdHTlyuZx9e49w7OgpADp3bcuyJes0HvNdmzcvp3Hj+uTNa0Zg4CX++GMxw4c7smjRLBQKHRITkxgxIvM9OxqXQ7rfskLrw74jIiLQ0dEhT548H6wLCAigfPmMu3G+ZNi3pmV22Lc2ZGbYt7aIB/Rlj+x8QN+XDvvWtC8Z9q1pXzLsWxs0Oew7/o++WdrP+Hf3DLdRKpU4Ojpibm6Oo6MjK1eu5Pbt2xj973uPI0aMoHjx4kiShJubG35+fujr6zN8+HBKliyZYflabyFZWFh8cl1mkpEgCILwBbTYQjp8+DCFChUi4c3XTYB+/fpRr169dNv5+fkRGhqKi4sL9+/fx9XVlXnzMv4+2g/1PSRBEIT/95TKrL0yEBERga+vLzY2Nhlue+XKFaytrZHJZJQtW5b4+HiioqIy3E8kJEEQhB+JloZ9b9q0ib59+yKTydIt37ZtGxMmTGDTpk2kpKi+1hIZGUnevG+7ny0sLIiMzHgE6g83l50gCML/a1n8Yqy3tzfe3t7q321tbbG1tQXg6tWr5M6dm5IlS3Lr1tsvlvfu3Zs8efKQmprK2rVr2bdvH926dfvoCOr3E9nHiIQkCILwI8niPaR3E9D77t69y5UrV/Dz8yM5OZmEhARcXFxwcFCNHNXV1aVZs2YcOKD6WoGFhQXh4W+nCouIiMDMzCzDOoiEJAiC8APRxpdce/fuTe/evQG4desWBw4cwMHBgaioKMzMzJAkCR8fH4oUUX0PrXbt2hw9epSGDRty//59jIyMREISBEH4f+cbfg/JxcWFmP/N+FKsWDGGDFHNXVijRg18fX1xcHBAT0+P4cMzN9u+SEiCIAg/Ei0npEqVKlGpUiUAZs6c+dFtZDIZgwcP/uKyRUISBEH4kXzHs31/FwkpPjUh4420RJmNzy9sH5h9b6y1Ohn392pTr4Rn2Ra7kXHxbIsdqvzwUSHfip48ez8O4lOSMt5IS4qbWmZbbI37jqcO+i4SkiAIgpA5kkhIgiAIQo4gEpIgCIKQI+SQZxtlhUhIgiAIPxLRQhIEQRByhO84IYnJVQVBEIQcQbSQBEEQfiBafuaqVomEJAiC8CP5jrvsfsiEJJfL8Tq1i9DnL+nXaxjLVs2jfsOfiImJBWD08Kncuhmg0ZiFChVgvetfWFrmQ6lU4rZxG6tWbWLqtNEMGNiL8HDVs0BmzXTmmNcpjcZ+o9egbnTq0w6ZTIbn1oNsc93FbxMHYW3XCElSEhkezewx8wh/EfHVsWT6upT9Zx4yPV1kOjpEH75AyOJt5OvfhnyDO2BQvADXq/YlLUr1N88/tDPmna1V+yp0MChdmBvVfyEt+usfW53L1IQ/l06nbPlSSJLElNFzSExIZLbzFIyMjXj29Dnjf5tOfFz8V8cCGLhwONWa1yIm4hUz7Mapl9v0b43NL61IS1Ny48RVdjm5U7FRVbpN7oNCV0FqSio7520h4KJ/lmOPcR5DHZs6REdEM7yFan4w+6n21LWtS2pKKiGPQ1gyYQnxMapjLV6+OKPmj8IolxGSUmJ0+9GkJKV83R8AKF6qKIvXv30CaJFiBVm+YB15zHLTvLU1SqVEZHgkU0bNIexF+GdK+nKFCxfA1XWJ+lzbuNGDlSvdMDPLzZYtKylWrDCPHwfTt+9woqNjNBoboESpYh8cu8uCdezbeYjF6+dRqEgBnj0NYezgKcS8itV4/Ax9xwlJJn0H7TurPBW+aPuhI/pTrXplcuUyUSekf4+e4uD+Y18cOy45MVPbWVrlw8oqP9ev3cLExJiz5w/wc88hdOnalri417gsW//FsSvkKZLpbUuVK8Gfq2fSv+1QUpNTcfFwxslxMZFhkcTHqb7933NQV0qUKY6T418ZlpeZmRrkRgYoXyeCQodye5x4OnM9UnIKaa/iKbPzDwLajlcnpHfltv2J/IM7cL/X9E+W3Ssx8zM1LFgxiyv/+bHLfR+6ugoMDA3Y9M9KnGYtw+eCL117d6Bw0YIsc1qTqfIymqmhbJ0KJMYnMnjxKHVCKl+/Eu1GdGWp/TxSk1PJZWFKbEQMRSuVICYsmuiXURQqW4Rxm39nfL2hnyw7o5kaKtepTMLrBMYvGa9OSDUa1+D6heso05QMnDIQALf5bsh15Cw/vJxFYxYRdCeIXHlyER8Tj/ITw4KDErOWOORyOaduHKJXq4G8io5VJ/6+g3tQqlxJZk90ylQ5j2JfZGo7K6v8WFnl59o1f0xMjLlw4SA9egyhX79uREVFs2jRaiZMGEaePLn5/ffMxS6WK3+mtnufXC7n9I3D9Gw1gN723XkVFcP65X/z66j+mObJxV9zV2SqnICXPlmK/zGvBn78ERIZye3mnfFGWvbDDWooUNAS25ZN2Lrln28a90VoGNevqR5cFRcXz927gRQoaPXN4hcvU4ybvrdJSkgiLS0N34vXaNq6sToZARgaGmi0f1n5WpWsZQodZAodkCDhVhDJwS8/u59Zx8ZE7jujkToYmxhTu14NdrnvAyAlJZXYmDhKlC6GzwVfAM6fuoRdu+YaiQdw7/Id4l+lb9k162PH4dV7SU1OBSA2QnVl/uRWENEvVY9ufnbvKbr6eij0st4x4X/Zn9jo9Ene76wfyjRVkgnwDSCvlepJnTWtaxJ0J4igO0GqOkXHfjIZfY161j/x9FEwz4ND07VCDY0MQQvXu6GhL7l2TdXKjIuLJyAgkIIFLWnXrgXu7rsBcHffTfv2LTUe+3313zl2m1ZN8NxxEADPHQexbd1U6/E/SktPjP0WvklCCgwMJDAwEIDg4GAOHjyIr6+vVmLNnT+FuTMWffBMEMfpYzhx3pPZ8xzR09PVSuw3ihYtRLVqFbnicw2Aob/9wn+XjrBqzQLy5DHVSswHAUHUqFuN3Gam6Bvq06B5PSwLqq76hk0ezMEr/9CqSwvWOm/QXFC5nPJHl1D12mZizl7j9bV7Ge4iM9DDtGlNoo9c1EgVihYvRFRENE7LZ+J5Yit/LvkdQyMD7t15gE2rJgC07mCLVSHtzlVmWbIAZepU4HfP+UzeMZviVUt9sE2t1vV4citInbS0oWXPllw5dQWAQiULATB3y1xcDrnQ7bduWonZplMLDu152/swesowTvgdoH3XVrgsWKuVmG8ULVqY6tUr4eNzjfz58xIaqroYCg19Sb58eTPY++u16dSSQ3u8ALDIZ07YS1V3eNjLCMzzZtN8kMosvnIArSekXbt24ebmhqurKx4eHmzYsIHExET27dvHnj17NBqrhV1TwsMiuXH9drrlf85eQqOf2tCqWXfMzHIzcsyvGo37LmNjI7ZuW83kSXOJjY3Ddf1WqlRqQv16bXgRGsY8p2laifso8DGbV3mwYvtiXLYu4v7tB6SlpgGweoEr7Wp34+ief+lh30VzQZVKAlqNxb/OIIyrl8WgXNEMd8nTog7xPnc0cu8IQEdHh4pVy+Hh9g+dmvfh9esEhjgMYOroOfSx784e7y0YmxiRkvz1900+R66jg7GpCX90msLOeVsYtnJcuvUFyxSmu2Nf/p6qvQ/oniN7kpaaxsm9J4H//W1qV8TZwZmJXSdS364+1RpW02hMXV0Fze2s8TpwXL1s2fzVNK/RngO7j9JnUHeNxnuXsbER27atYeLEOcTGaub99CXeHPvRd449J5CUUpZeOYHWE9J///3H3LlzmT17Nl5eXkycOJFu3boxbdo0Lly48Mn9vL29cXR0xNHRMdOxfqpbg5atm+Fzw5s1G/6ioXVdVqxdwMsXYQAkJ6ewfeseatSs8tXH9TEKhYKtHqvZsX0f+/eprppevgxHqVQiSRJuG7dRu5ZmPxDetX/bIfrZDWZol1HERMfwNCg43fqje71p3qaJxuOmxcQTe/Empk1rZritWYfGRO4/q7HYoSEvCX3+khu+qu5SrwPHqVS1PA8DH2PfYyRdbPtxcI8XTx9pd/bwqNAIrnpdAiDoeiCSUiKXuao1bGZlzsi1k3Adt5ywJ5m7T/KlbLrZUMemDs4Ozupl4SHh3Lx0k5ioGJISk7hy8gqlK5fWaNzGNg24fTOAiLDID9Yd2uNFy7aa6yp9l0KhYNu2NezY4cm+fUcB1blmZaXqFbCyyk9YmGYHU7zv/WOPCIskX34LAPLltyAyPEqr8T9JdNl9mo6ODnK5HH19fSwtLTEyMgJAT08PmUz2yf1sbW1xcnLCySlzNyUB5s1ZQs1Kzfipqi2/DRrP+TOXGDl0Mvkt86m3adXWloA797N+QJ+xavUC7t4NZMXyt91illZvY7fvYMft2xl3a2WVmUUeVcxC+WnWxhovT2+KlCisXm9t15BHgU80EkthboqOqTHwv264xtVIDAz+7D7yXEaY1KvEq/99cGtC+MsIQp+/oESpYgDUb1yHwLsP1d0lMpmM4eMGse3v3RqL+TF+x3yoUL8yAJYlCqDQVRAbGYOhqRFj3Kaye+FWAq/e1UrsWk1q0X1Yd2YPmk1S4ttHOPie8aVE+RLoG+gj15FTuV5lntzXzP//jbadW6brritW4u1AnGZ21jwMfKTReG+sWbOQu3cDcXFxVS87dMibvn27AtC3b1cOHvxXK7HfaNvZLt2xn/A6Q6ee7QDo1LMdx4+e1mr8T/qOu+y0PuxboVCQlJSEvr5+uuTy+vVr5PJvM6Zi1fqFWFiYI5PJ8L95h0njZms8Rv36tendpwv+NwO48N8hQDXEu3v3DlStWgFJgsdPgnEYNVXjsd9Y4DqX3Ga5SU1JZeHUJcS+iuP3RZMpVqoISqVE6LNQ5k/OeIRdZujmN6PYkjHIdOQglxF14Dwxx6+Qb2A7LId1RjefGRX+dSHmxFWeTFKNNMrTqh4xZ66hTNDsc2/mTnFm0Zq56OrqEvz4GY4Os+nUsy197FXdRf8eOsluj/0aizfUZQzl6lXCxCwXiy6uZd+SHZzdeQL7hcOZ47WYtJRUXMerjtnml9bkL2ZFe4dutHdQ3cP5q99c9aCHLzVp+SSq1q+KqZkpmy9txn2xOz1G9EBXT5c/t/4JwF2/u6yYuoK4V3Hsdd3L0oNLkSSJKyev4HNCc6O5DAz1adCkLjMnzFcvGzd9BCVKFUMpKXn+NJRZmRxh9yUaNKhNnz5duXnzDv/9dxiAmTOdWbRoFe7uq+jfvydPnz6nT59hGo/9hoGhPg2b1GHmhLfDv9e7/M2S9fPp2qcDIcEvGDM48707mpRTut+yQuvDvlNSUtDV/XAQQUxMDNHR0RQtmvF9hy8d9q1JmR32rQ1fMuxb07L9AX1fMOxb0/6/PqAvq8O+NSWzw761IavDvjVFk8O+o7o2zdJ+ZrtPaawOWaX1FtLHkhGAqakppqbaGXEmCILw/9X33EL6IWdqEARB+H8rh9wPygqRkARBEH4gkkhIgiAIQo4gEpIgCIKQE3zPLaQfbi47QRAE4fskWkiCIAg/ku+4hfTJhLRq1apMFTB8+HCNVUYQBEH4Ot9zl90nE5K5ufm3rIcgCIKgAT9kQurVq9e3rMdnGSkMsi32q6Ts++Z8OV3tT5//Kd3ig7ItNsCJwtl37PUe3cq22La5s29Wkle6JtkWGyBMNzrbYgfFhGZbbE37IRPS+/z9/blw4QLR0dFMmjSJhw8fkpiYSMWKFbVZP0EQBOFLSJ+etDqny9QoOy8vL9asWYOFhQW3bqmuHlXTv2/TauUEQRCELyMps/bKCTLVQjp48CDTp0/H0tKSgwdVj+gtXLgwz55l3wSYgiAIwock5ffbQspUQkpISCBfvnzplqWlpaFQiFHjgiAIOUlOae1kRaa67MqXL8/+/emfJ+Pl5SXuHwmCIOQwkiTL0isnyFQTx97eHicnJ44fP05iYiLjxo1DoVAwZcoUbddPEARB+ALfcwspUwnJ3NwcJycn7t+/T1hYGHnz5qVs5k41/QAAIABJREFU2bLf7ImvgiAIQub88PeQ3pDL5ejp6aGjo6Ot+giCIAhfQbvPANeuTCWkJ0+esGjRIhISEjA3NycyMhJDQ0MmTJiQqUeQC4IgCN/GD99CWr16Nc2bN6dDhw7I5XIkSWL//v2sXr2a+fPna7uOXySXqQlOS2dStkIpJElissNs7No1x8bOmpTkFB4/CmbSqJnExsRpPHbhwgVwdV2CpWU+lEolGzd6sHKlG2ZmudmyZSXFihXm8eNg+vYdTnR0zFfH+9V5BNWb1yYm4hVTWo4BoMuYnjT92ZbYCFX5O523cv2kLyWrlWbQ/GGqHWUy9i7dwRWvS19dhze+5d9dpqeL1cbFyHR1QaHDa++zRK/eTN45E9GvVQUpTjW7RvgMZ5LvPsC4TXNyD+gJgDIhgYg/XUi59/Cr6/GGXC7H69QuQp+/pF+vYerlfy6cRq/enSlVuLbGYn3J/7xyo2r0dOyLQldBakoq2+b9ze0L/lmOPeWviTS0rUdUeDT9bAYBkCtPLuauno5VEStCn4Yy/bc5xL6Ko2ipIkxbMomylcuwbsFGtq3d+fUH/z/6+nocOOqBnp4eCoUOB/Z5sWCey/+xd99hURwNHMe/HEcHRZCi2FssiRg7CCiKDcVeY4saG2rUJCqxd8UeK/ZYscXeUFRU7CUoKqJYYkFAitLb3b1/oBc1Koh7Ut75PM89j+7u7W/2jru5mZ3dYf+RLRgbGwFgYWHGtauB9PxB+nttrlgxF9fmjXjxIorqNVwA2LRxGRUqlAGgoGkBXr2MpXadZpJnZ0VerpC0VKrMG3i9evVi3bp175wzUigU9OnTh/Xr12u0gABlCn+f5W3nLJnC5Qt/s33TbnR05Ogb6GNb/VvOn7mMQqFg9ISfAfCcsihL+3ueEJ3lbGtrS6ytLQkIuImxsRHnzh2gU6f+9OjRgZiYl8ydu5zffhuEqWlBxo2blen+2lvW+OT6b2pXJiUxmQHzf37nyyk5MZlDK/e+s62uvi7paekoFUpMLQsx/fB8htbui1Lx4TOg5z7z1kFSv+6Z3TpIy0AfVVIyyLUpsm4B0bOXY9KxJYmnL5Doe+adbfVsK5P24DHKuHgM6tXCdGAPnvf4+aP7rvvoaZbK+MaAwb2wrfYtJibG6grJtloVfhrUE9cWjT6rQsrs1kGf856XrFKaVy9e8jIihmIVSjBq43h+rtPvo/t+mP7pW/fY1qlKUkIS4//wUFdI7mP7E/syjk1Lvek+uCsmBY1ZPmMVpuamWBezwqlZPeJexmepQgqOy/rrbmRkSEJCInK5nINHvRkzehpXL19Xr1+3cTGHDx1nu/eeLO0vLjUpy9kODnWIj09g7ZqF6grpbZ6zxvMqNpYZM/7I8j5Tkp9kedvMPLRtnK3nlb5+TLIyZFeWRiXY2tpy7dq1d5b9/fffVKtWLVuhS5YsydbzMmNsbERtu+ps37QbgLS0dOJi4/H3u4BCoQDg7yuBWBe10kh+WFgEAQEZv0Dj4xO4cyeEokWtaNmyMZs2/QXApk1/4ebWRJK84Eu3iX8Zl6VtU5NT1ZWPjp6OpB3NOfG6q5KSAdCSy0Eu51O/q1Ku30YZl9EyS7kRhLaVxUe3/VxFilrh0qQ+mzfuVC+TyWRMmDqSqRPmSpbzxue85//cesjLiBgAnt59jI6eLnLd7F87eP3iDWLfa9k7Nq3H4R0+ABze4YNTMwcAXka95M71YNLTFNnO+5SEhIxWsI6OHJ333n9jYyMcnepy6IBmvmD9/S8SE/Pxyrt9h5Zs37b3o+s1TaXUytYjN8jS9BMymYz58+dTvnx5zM3NiYqKIiQkhFq1amUa4Onp+c7/VSoVt27dIiEhAYDRo0dnt+z/UbyUDdFRMcxePJlKVSpw80YQU8bMJikxWb1Nx26tObDnqGSZH1OiRDGqVavC5csBWFoWJiwsAsiotCwsNHvj0MY9m+PQrj4PA++zeeqfJMZmvNZlq5Wn35zBFLaxwGvEoo+2jj5XjrzuMhlFvZchL16UuG37SL15Bzq5UWhIb0z7dyfp0t/E/LEG0tLeeZpx22Yk+V+WrBhTZ/7O1AlzMTYxUi/r078bPodPEhH+QrKczHzsPX+jlqsd/9x6QHpquqS5hQoXIioioxchKiIaU3NTSff/MTKZjOOnd1O6TAnWrtrMtSs31Otc3Rpz+tR54uMSPrEHzXBwqENEeCQh9x999ez84KMtJDMzM/XD2tqaVq1aUalSJSwtLalUqRJubm5YW1tnGvBmAETLli1xc3PDzc0NAwMD9b8/xtfXFw8PDzw8PLJ8MHK5nCpVK7J53Q7cGnYlMSGJgT/3Ua93H9GX9HQFe3ccyvI+s8PIyBBvby9GjpxCXJz056o+xXfTEX5xcmds8195GRFDt/E/qtfdD7iHR+PhTGg1Cjf3dhktJQnkyOuuVBLaeSBPm3ZF99tv0ClbiphFa3jWpg+h3YagXdCEgr07v/MU/Zq2GLdpTswfqyQpQuOmDYh8Ec2N67fVy6ysLXBr3ZQ1KzZJkpEVn3rPAWzKF6eLRw/W/u711cqkaUqlEmeH1lSt5ET1GlWpWKm8el27Di3ZtfNAjpSrc6fWbN+ec60jyKcXxko1/cTMmTM5dOgQu3btokePHpQqVQpdXd1M7/Lg4uKCi0tG/+z21T5ZynoeGk5YaATXr2V0mx3Z78vAYb0BaNfZjYZNnOjebsAXHE3mMm4668W2bXvYu/cIABERkVhbWxIWFoG1tSUvXkRqLD828pX63ye9j/Hr2rH/2SY05BkpSckUq1CCh4H3vzgzJ193ZVwCyVeuY1CvJrEbXnebpaURv9eHAj07qrfTKV8a84m/ED54DMpXWevyykytOt/TpLkzjZo4oaeni7GJMacu7Cc1JZULf2f8zRoYGnD+2hHsqmvuBPen3nMza3OGrxyN1y+LiHgcLnl2TGQM5pZmREVEY25pxsuorzuFROyrOM76X6KRiyN3gu5RyMyU6jW+o5cGBjNkRltbm9atm2Fn7/rVs9+Wly+MzfKVrenp6Tx9+pSgoCBu376tfmQaIJPRsmVL3N3d2bVrF2vWrFGfV5BaZEQUz5+FUbpcSQDsnWpzL/gBTg3tGfDzj/TvPpzkpORM9vJlvLxmExwcwqJFq9XLDh70pXv39gB0796eAxrq2wYwtSyk/nfNpnV4GvwYAIvilsi0M95ucxsLipSx4cXTCEkyv/brLitUENnrLjItPV0M6lQn7eETtAv/O6mkoXM90kIeAaBtbYHlvIlEjvMk/bF0NwSeMWUB1as4U6uqCwP7/srZ0xepWKouVb9xolZVF2pVdSEpMUmjlRF8/D03LGDIr+vGsn32Ju5duaORbP+j52jesSkAzTs25YzPWY3kvM3cvBAFCpoAoK+vh1MDe+7dyxg12bpNM44e8SMlJVXj5Xhfo4aOBN+9z7NnOTu3klKlla1HbpClM5x3795l/vz5JCUlkZKSgp6eHqmpqZiamrJ8+fIsBZmbm/PLL79w7do1DAwMvqjQnzLpd08Wes1AR0fO43+eMWroRPYc24Suni4bdmaUNeBqION+my55tr19Tbp1a09gYBAXLmR0T02cOIe5c5exadMyevXqzJMnoXTrNiiTPWXN4EUjqGT3LcaFTFh0YRV/LdhKpbpVKFm5NCqVisinL1g7JqObpkLNSri5t0WRpkClUvHnuJXEx0jTUoCv+7prFzaj8NRRaMlkINMi4ehpks5cxGrlbLQLmYIWpAbfJ2paxign0/49kJkWwHxMxsg6VbqC590Gf3E5csLnvOeNe7liVcqaNkM70mZoRmvRs8cUYqNefSrioyYtHcf3draYmhVk95VtrJn7JxuXejPVawItuzYn/FkE4wZMBsDMohBrDnthZGyIUqmiU7/2dGvQm8T4L5/w0srakiVenmhry5DJZOzdfZijR/wAaNu+BX8sWPnFGZ+yYcMSnBzrUriwGfdDLjF12jz+/HMbHTu1ytHBDG/klu637MjSsO8xY8ZQt25d3Nzc6NOnD+vWrWP79u0YGhrSsmVLjRfyc4Z9S+1zhn1LLbNh35r0ucO+pZazM8Z+3rBvKeXkjLGZDfvWtM8Z9i21zxn2rQlSDvu+UyF7XYYV72r23HpWZKnLLjQ0lJYtW6Kl9W/N265dO/XcSIIgCELuoFJl75EbZKnLzsDAgOTkZAwNDTE1NeXZs2cYGxuTlJSzvyoEQRCEd+WWa4qyI0sVUq1atbh69SqOjo40aNCAyZMno62tTZ06dTRdPkEQBOEz5JYBCtmR5fmQ3mjdujXlypUjKSmJ6tWra6xggiAIwufLy4MasjWhUZUqVahWrRpTp06VujyCIAjCF8j355A+RKVSZek6JEEQBOHryfdddoIgCELekJe77ESFJAiCkI/klu637PhkhbRz586PrktPl/auwZ8Sl/blV3dnl0KpmdscZcWtFOnvPZZV/Yyq5Fg2gPOTwBzLvljGJsey3UKf51i2ibbm7qCSFUnpX/92P2/oaUtzo+HcIN922T1//ukPh4ODg6SFEQRBEL5Mvu2yGzp06NcqhyAIgiCBfNtCEgRBEASA1NRUJk6cSHp6OgqFgrp169KpUyciIiJYuHAh8fHxlC5dmqFDhyKXy0lLS2PJkiU8ePAAExMThg8fjqWl5SczsnUdkiAIgpA7qbL5yIyOjg4TJ05kzpw5zJ49m4CAAO7evcumTZto0aIFixYtwsjIiBMnTgBw4sQJjIyMWLx4MS1atGDz5s2ZZogKSRAEIR/R1HxIWlpa6OvrA6BQKFAoFGhpaXHr1i3q1q0LQIMGDbh8+TIAV65coUGDBgDUrVuXmzdvktnkEqLLThAEIR/R5KAGpVLJ6NGjCQsLo2nTplhZWWFoaIi2tjYAZmZmREdnTNkTHR2Nubk5kDGbrqGhIXFxcRQoUOCj+89yhXTz5k3OnTvHy5cvGTVqFA8ePCA5OTnTqcgFQRCErye7M5j7+vri6+ur/r+LiwsuLi7vbCOTyZgzZw4JCQnMnTuXZ88+PgPzh1pDb09h9CFZqpB8fHzYv38/zs7OnD2bMUWxXC7H29tb3M9OEAQhF1GRvRbShyqgjzEyMqJy5crcu3ePxMREFAoF2traREdHY2ZmBmTMEh4VFYW5uTkKhYLExESMjY0/ud8snUM6cOAA48ePp3379shkGU8pVqzYJ2tHQRAE4etTqrL3yExsbCwJCQlAxoi7wMBAbGxsqFKlChcuXADAz8+PmjVrAlCjRg38/PwAuHDhAlWqVJGmhZSUlISFhcU7yxQKBXJ57jwFJZPJ8D31F89Dw+nWeSB9+3VjgHsvSpcpyTel6xIdHaOR3JUr5uLq6sKLF5F8Xz3jl8akib/h5tYUpVJJxItIfvrpF54/18wdGLr170zbbm6oVCpCgu4zcfgMUlNSGezRn8ZuzigUSnau3433mo/fgSOrTIqY4bZgIEYWBVEpVQRsOcmVdT5UdK2Nw4h2FC5XlD9bTSQsMGMq9CK2ZWg+s2/Gk7XAf+Fu7vpc+eJyAJz5+xDx8YkoFQrSFQpaN/qBSlUqMG3eOAyNDHn2OJThA38nPi7hy8N0dbBevQAtXR3Q1ibx+GleeW3AfNJI9GtURRmfkRE5cQ5pd+//+7TK32C9fhGRHtNIPH7my8vxWvf+XWjXzQ1UKu4F3Wf88OlYWJkz22sqBUwLEBQYzJghk0lP+/I7q/w+byT1XOoSE/mSHo0y3ksTUxOmLh+PdXFrwp6EMX7gFOJexVOibHHGLhhFhW/Ls9JzLd4rtn9x/hs2NkVYtXo+VlYWKJVK1q31Ztmyder1w4b1Y8bMsZQo/j1RUdJ+1m1sirBi1Vx19p/rtrJ82Z98+11FFv4xDSNjIx7/85Sf+owgLi5e0uysUmazhZSZmJgYli5dilKpRKVSYWdnR40aNShWrBgLFy5k69atlC5dmoYNGwLQsGFDlixZwtChQzE2Nmb48OGZZmipMhv2AMydO5dy5crRpk0bevfuzbp16zhw4AAhISFZCvlSFgW/+aztBw7+kWrff4uJiTHdOg/ku6qVePkylj0HNtC4QYfPqpBeJWf9S8zBoQ7x8QmsW7tQXSGZmBir/zAHD+5DpUrlGTLk9yztr0qhklnOtrAuzLp9y2nv1I2U5FQ8V07B//gFtLSgVr3qTPh5OiqVikKFTYmJfJnp/jrplf7keiNLU4wtTQm/+QhdI316H5jKzv4LQAUqpYpmM/pwYvoWdYUk19dFkZaOSqHEyNKUvoens7j2UFSKD/d4r4zP+q2Dzvx9iFaNfiAm+t/j2uO7mZkT5nPx3FU6/tCG4iVtmD9zaZb251fS4pPrtQz0USUlg1wb6zULiZ67DJP2LUk6c+HDlY1MhtVyT1QpqcTvPfLJCsktNOt/b5bWFqzf50Ubpx9ISU5hzsppnDl+DsdG9hw/6MeRvb6M8xzF3dv32L5+d6b7y+zWQbZ1qpKUkMT4PzzUFZL72P7Evoxj01Jvug/uiklBY5bPWIWpuSnWxaxwalaPuJfxWaqQAmIeZOm4ra0tsLa2JCDgFsbGRvif3U+Xzv25cycEG5siLFvmSYVvyuBQzy3LFZK2VtYGHFu9zr7+Ovu0/z66dhnAipVzGTtmBmf9L9G9Z0dKlSzGtKkLsrRPgNiErB17Vhy36pyt5zUK3yZZGbIrS+9Cnz59OHfuHEOHDiU5OZlffvmF06dP06tXr88OvHPnDgcOHOD69euf/dysKFLUisZNG7Bpw7+tgMAbQTx5rPnuRX//i8TEvPtl//avJCNDg0yHPX4JbW1t9PT10NbWRt9AnxdhkXTs1ZaV89apc7NSGWVFQsRLwm8+AiA1IZnIkFBMrMyICgkl+sF/bzmVnpyqrnzkejpZu/DhC5QpV4qL564C4O93nmZujSTbtyopGQAtuRzk8kzvZmnSpQ0Jx8+giJbmtX/b++95ZHgUtevV4NiBkwDs234I52ZOkmRdv3iD2Jex7yxzbFqPwzt8ADi8wwenZhm3E3sZ9ZI714NJT5P+XpBhYS8ICLgFQHx8AsHB9yla1BoAz9njGTdupsZuMBoe9oLr72SHULSoNeXKl+as/yUATh73p1XrZpopQBYos/nIDbJUIZmZmTFr1iwGDx7M4MGD6d+/P7NmzaJQoUKZPvf33/9tDfj6+rJmzRqSkpLYuXMne/bsyX7JP2L6rDFMnjAHpTK3vMQwZfIo7odcomvXtkyePFcjGS/CItmw3JvDV3dx7MZe4mMTuHDqEsVK2tCkdSM2+6xhyZa5lChdTPLsgsUKY1WlJKEB9z+5XdFqZfnp2Cx+8pnJkbHrPto6+lwqFWzY6cW+49507dkegLtBITRu3gAA19ZNKGJjLUkWADIZRby9KOa7k+SLV0m9eQcA08F9KLJtJYV+HQQ6GTfr1LYwx9C5HvE7D0iX/1pE2AvWL9/C0au7OX5jP/Gx8dy+cYe42HgUioyKIPx5BFZFPt3i+xKFChciKiJjmG9URDSm5qYay/qQEiWKYWtbmcuXA3Bt4cLz0HACA4O+UrYNVW2rcOVyAEG37+LaIqNXpE07V2yKFfkqZfgQFVrZeuQGWb4wViaTUbFiRRwcHKhYsaJ6cENm3nwwAI4fP8748ePp2LEjY8eO5cyZj3dd+Pr64uHhgYeHR1aLSOOmDXjxIpobr3/B5BYTJs6mbLnaeHvvxn1Qb41kmBQ0oUEzR1rW7kgT29YYGOrj2r4Juno6pKak0q1pX3Zt2s/EBWMkzdUx1KOt1zB8p2wiNT7pk9uGBtxndWMP/mw1ATt3N7T1pLnDcgfXXrg17ELvzoPp0bczte2qM+rnifTo24V9x70xMjYkLTVNkiwAlEqedx3I02Zd0KtSEZ2ypXi5ZA2h7XrzvPtgZAVMKPhjRrdJod/ciVm0GjTwA8mkoAnOzRxpXrs9LrZuGBjq49DQ7j/b5eXpCD7FyMiQLd7LGTVqCunp6YwaNYSpU+d/teyNW5bhMWoqcXHxuA8aTf8BPTjlvxcTYyNp/94+U15uIWVpVMLgwYM/OjpiyZIln3yuSqUiPj4elUqFSqVSXxSlr6+vvpjqQ94egrhmeeb93wB16lanWfOGuDR2Ql9fD2MTY5atnIN7/5FZer6mbd22h7171jNl6jzJ913HqSahj0OJicroFjpx6BS2tb4jPPQFvgf91MsmLZSuQpLJtWnnNYxbe85x90jWByhEhYSSlpSCRYVi6nNMXyIi7EXGfiOj8Tl4Atvq37Jq6QZ6dhgIQOmyJWnYRJpuq7ep4hNIvnodA/taxG7ckbEwLY34fT4U6NkRAL3KFbCYORYAmWlBDBxqo1IoSPI798X5dZ1q8fTxc/V7fvzQKarV+g6TAsZoa2ujUCiwKmKpfn00ISYyBnNLM6IiojG3NONllPTdkh8il8vZssWLbVv3sG+vD1WqfEOpksW4cPEwADY21pw9d4D6Tm0ID5f2+OVyOZu2LGP7tn3s35fRXXnv7gPatMo4hVGuXGmaNnOWNPNz5JbKJTuyVCENHDjwnf/HxMRw5MgR6tWrl+lzExMT8fDwQKVSoaWlxcuXLzE1NSU5OVny8ynTJs9n2uSMX0j2DrUZPLRPjldG5cqVJiQk40u3ZcsmBAd/ulsru8KehvNdjW/RN9AjOSmF2o41uX39DglxCdR2qMFe74PUsP+exw+eSJbpOvsnokJCubz6cKbbFixuQWxoFCqFkgI25piVKcKrp1/+RWFgaIBMpkVCfCIGhgY4OtuxaM4KzAubERUZjZaWFkN+7cfmdTu+OAsyKhVVejqq+AS09HTRr1Od2D+3ol3YDEVkRteVobM9aSGPAHjm1kP9XPNJI0k6c0GSyggg7GkYVWtUUb/ndRxrcut6EKbnrtG4pTNH9vrSqpMrfj7Sjep7n//RczTv2JRNS71p3rEpZ3zOaizrbcuXexIcHMLixWsAuHUrmFKlaqrX3w7yx9Eh64MaPsfS5bMIDr7P0tfZAIUtzIl8EYWWlhYjRw9mzZotkudmVW7pfsuOLFVI33333QeXzZw5kxYtWnzyuUuXfnhkk5aWFiNHfp3Kot+AHgwZ9hOWVoU5dW4fvsdOMWLoOMlzNm5YgpOTHYULm/Hg/mWmTJ1H82YNqVChDEqlisePnzI4iyPsPtfNv2/je+AkW46uQ6FQcCfwLn9t3Iuevh4zlk2kW//OJCUkMeWXWZLkFatZge/aOxIR9Jg+h6YDcGrOdrR1dWg8uSeGZiZ0Wvcb4bf/YVvP2RSvWYG67m4o0xSoVCp8xv1JUsyXD4stbGHGig0Zo5m05XL2/XWI0yfO8WP/H+jZtwsARw4eZ8cWac5XaluYUXjyaNCWgZYWicdOkXTmIlYr5iAzNQUtSL17n+jpCyXJ+5TA1+/5tqPrUSjSCQq8y86Nezl97ByzV0xliMcA7ty8y64t+yXJm7R0HN/b2WJqVpDdV7axZu6fbFzqzVSvCbTs2pzwZxGMGzAZADOLQqw57IWRsSFKpYpO/drTrUFvEuO/fLJNO7ua/NCtPTcDgzh/4VBG2SbOxsfH74v3nZm6djXp+kM7bt68g//5jPOCUybNpWzZUvTrn/HjY98+HzZtkOYHUHYo8259lLVh3x+SkJCAu7s769evl7pM//G5w76l9DnDvqX2OcO+pZbZsG9N+5xh31LLbNi3Jn3OsG+p5fSMsVkd9q0JWR32rSlSDvvea/1Dtp7XOiznWnVvZKmF9P5U5ikpKVy7dg1bW1uNFEoQBEHInrw8hiVLFdL7U5nr6enRtGlT9a3FBUEQBOFLZVohKZVKqlatip2dHbq6ul+jTIIgCEI25eVRdpl2nMpkMtauXSsqI0EQhDxAqaWVrUdukKUzedWrV+fatWuaLosgCILwhTQ1hfnXkKVzSCqVinnz5lGxYkX1DIBvuLu7a6RggiAIwufLy112WaqQrK2tcXNz03RZBEEQhC+Ul69D+mSF5O/vj4ODA126dPla5REEQRC+gKbmQ/oaPnkOadWqVV+rHIIgCIIE8u05JE3O3fM5EtNScizbzMAkx7LjFck5lr037WmOZQPoaefcqM5qdx/lWHbYxRU5lm1jNyjHsgGUOfh9o1RJP29TTsm3XXZKpZKbN29+cgfffvutpAUSBEEQsi/fDmpIS0vDy8vroy0lLS2tTKefEARBEL6e3NGvlT2frJD09fVFhSMIgpCH5NsuO0EQBCFvybdddrllUIMgCIKQNfm2QtqwYcPXKocgCIIgAZXoshMEQRByg3zbQhIEQRDyFlEh5RI2NkVYtXo+VlYWKJVK1q31Ztmyder1w4b1Y8bMsZQo/j1RUTEaKYNMJsPHbwdhoRH06JJxoaHHuGG4tWmGQqFg/dqtrFmxSSPZJgWMmbFwPOUrlgOVCo9hk/lxwA+UKVfy9XoT4mLjaOWcvSmOP6VT3/a07tYSLS3Yu/kg21bvZJrXBEqULaEuW1xsPD0b/yRpbumyJZm/aob6/8VLFmWR50qsiljg3MSRtLQ0Hj96ypifpxAXGy9pNoCeni4HfbzR09NFWy5n354jzJr+B4uWzuT76t+ipaVFSMgjBg8YRUJCoiSZzQaMw9BAH22ZDG1tGVvneHDn4ROmenmTmpaOtraMsf278F35Ujx8Gsb4JRsJevCEoT+48WObxpKUQU9Pl/1HtqCrq4tcrs3+vT54zljE/iNbMDY2AsDCwoxrVwPp+YO0N2AuVqwIq1cvUH/O167dwtKl65gxYwyuro1ITU3j4cN/6N9/JK9exX6V7HbtXBk7dgQVK5bD0bEV164FSpr7OfLymf98VSEpFOmM+X0aAQG3MDY2wv/sfk6cOMOdOyHY2BShYUNHHj/W7B0I+g3qwb3gB5iYGAPQpVtbbIoVwaGWKyqVisKFzTSWPW7GSE6fOM/QPqPR0ZEKLrZNAAAgAElEQVSjb6DP8H6/q9d7TB5BvAa+lMt8U5rW3VrSp8VA0lPTWbhlNueOn2fcwCnqbX6eMIj4uATJsx/e/4e2DbsBGT8GTt04hO+hk5QuV5L505aiUCj4dfwQ+g/7kXlTpb+EISUlldYtepCQkIhcLufwsa34Hj3FWI/pxMVlvNbTZo6h34AeLJwv3V0Y1kwZTqECxur/L9iwm4GdW+BYvQpnrt5kwYbdrJ06ggLGRnj07ciJS9cly4aM427bsqf6uA8e9cb32Cncmv37Y2fdxsUcPnRc0lyA9HQFHh7TCAi4ibGxEefOHeD4cX+OHz/D+PGeKBQKpk3zYORId8aNm/VVsm/dukuXLgNYsmRG5jsRPipL8yF9iXv37pGYmPHLMDU1le3btzNr1iw2bdqkXi6VsLAXBATcAiA+PoHg4PsULWoNgOfs8YwbNxNNDhwsUtQKlyb12bxxp3pZrz5dmOe5TD1iMTIyWiPZxsZG1Kr7PTs27QEgLS39Py0C19Yu7N99RPLsUuVLcOvabVKSUlAoFFw7H0D95o7vbNOolTPH9kj/5fQ2O6daPHn0lNCnYZz1u4hCkXE7mOtXb2Jd1EpjuW9aPjo6cnR0dFCpVOrKCMDAQE/jI1a1tLRISEwCIC4xCQuzggCYm5rwbflSyLW1Jc9857jl8neO0djYCEenuhw6cEzy3LCwCAICMu4gEx+fwJ07IRQtasXx42fU7/mlS39jY1Pkq2UHB4dw794DyfOyQ6mVvUduoPEKafny5ejp6QGwbt06EhMTadOmDXp6eixbtkxjuSVKFMPWtjKXLwfg2sKF56HhBAYGaSwPYOrM35k6YS4q5b+9uCVLl6B1u+b4nNzBlh0rKF2mpEayi5eyIToqBs/Fk9h7YjPTF4zHwFBfvb6W3fdEvojmnwdPJM9+cOch1epUpUChAugZ6GHfsC5WRS3V66vVqUr0ixiePHwmefbbXNs04eAun/8sb9+1FaePn9NYrkwm4/S5fdx9eBG/E/5cvZLRGlmyfBbBDy5QvkJZVnpJOGJVS4sBkxfT+beZ7DzqD8CoPh2Yv2E3jfuNYf76XQzr1lq6vI+QyWSc9N9L0P3z+J08y7UrN9TrXN0ac/rUeY20it9WokQxqlWrwuXLAe8s79mzEz4+fjmSndOU2XzkBhqvkFQqFdqvf509ePCAH3/8kYoVK9KxY0fCw8M1kmlkZMgW7+WMGjWF9PR0Ro0awtSp8zWS9Ubjpg2IfBHNjeu331mup6tDSkoKTZ07smnDThYsmaaRfG1tbapUrciWdTtp3bAbSYlJDPi5t3p9y7bNOPCBL2spPAp5zMZl3izeOpeFm2dz7/Z90tP/vVllkzaNNN460tGR07CpE0f2v5szYHhv0hXp7N95WGPZSqUSJ/tWVPnGgeo1balUuTwAQwZ5UKmcPXeD79O2fQvJ8jbM+JXt835n2bghbD18iiu37rH9yBlG9u7AsVUzGNm7AxOXaeY85duUSiXODq2pWsmJ6jWqUrFSefW6dh1asmvnAY3mGxkZ4u3txciRU95pkY4aNQSFIp2tW3d/9ezcQFRIn1C8eHFOnjwJQMmSJbl//z4AoaGhyOUfP4Xl6+uLh4cHHh4en5Unl8vZssWLbVv3sG+vD2XKlKRUyWJcuHiY20H+2NhYc/bcAaysLLJ/UB9Qq873NGnuzOUbvnitmUc9pzosWeFJaGg4B/cdBeDQ/mNUrvKNpLlvhD2PICw0guvXMroTjuz3pUrVikBGZdWkhTOH9hzVSDbAfu9D9Gran0HthhH7MpanD5+qsxu4OnJs30mNZQM4NrLnduAdol782yXapnMLnJs4MHLQeI1mvxH7Kg7/Mxdp5OKkXqZUKtn110FatW4qWY6lmSmQ0R3XsI4tN+89Yp/fBVzqVgOgiX11bt77R7K8zMS+iuOs/yUauWR00xYyM6V6je84psEWilwux9vbi23b9rB377/d0N26tcfVtRE//jjsq2fnFnl5+gmNV0gDBw7k9u3bDB06lKdPnzJu3DiGDBnCihUrGDBgwEef5+LiwqxZs5g16/NOSi5f7klwcAiLF68B4NatYEqVqknlSg5UruTAs2dh1LNvSXj4iy86rvfNmLKA6lWcqVXVhYF9f+Xs6YsMGTCaIweP4+BUFwB7h1o8uP9I0tw3IiOieB4aTumyGV2Cdo61CQnO6NO2r1+bByGPCHseoZFsgELmGV+SVjaWNHB14ujrFlEtxxo8CnnMi+fSvt7va9G2KQd3/VvhOjjb8dOQngzq8SvJSZqbvsS8sBkFCmZMUaKvr0cDZ3tC7j18p2u2WfOG3L0rzfmFxOQUEpKS1f8+fz2IciWKYlGoIFdu3QPgYmAwJYpI+4Prfebmhd45bqcG9upzKK3bNOPoET9SUlI1lu/lNZvg4BAWLVqtXta4cX1+/XUQHTr0JSlJc1O3fCg7N8nL55A0PsrO0NCQwYMHk5SURHh4OEqlEjMzM0xNTSXPsrOryQ/d2nMzMIjzFw4BMGnibI33JX/K4oWrWLZyDv0H9SIhIZFfftbcr/Wpv89mntc0dHR0ePLPMzx+ngRAy7ZNNdZd98bM1VMoWKgA6WnpzB2zkLhXGd0YjVs35NieExrN1jfQo1792kz87d8RTuNnjURXV5e1O5YCcP1qIJNGSjviCsDayoJlK+egrS1DJpOxe9chfI6c5PDRrZgUMEZLS4ubgUH8OnyiJHnRL+MY7pkxWk+hVNLcsSYO1atgaKCH55odKBRKdHV1mDgoY+RhZMwruoz0JCEpGZmWFpsOnGTPovEYGxp8UTmsrC1Z4uWpPu69uw9z9IgfAG3bt+CPBSu/aP+fYm9fk27d2hMYGMSF15/ziRPnMG/eJPT0dDlwIKO78tKlv/n557FfJVtPT5f58ydTuLAZu3at48aN27Rq1VPS7KzKLd1v2aGlygM3rDMyLJVj2Sa6X/bB/aJsHcMcyzbXybmJCQFepmv2ZPinhCdq5hq1rPh/nqAvIQcn4sxpSUnSdbHOLNk9W8/7/R/Nn3fMTL66DkkQBOH/nTLXnBH6fKJCEgRByEfycpedqJAEQRDykbzbPhIVkiAIQr4iWkiCIAhCrpBbhnBnh6iQBEEQ8hExqEEQBEHIFfJudSQqJEEQhHxFnEMSBEEQcgXRZZePpSrScyw7Raa5e4FlJk6WlGPZAGlKReYbaYi1oeYmUcxMqXpDcyz7ya5fciwbwLrNnBzLNtbRz3wjQeNEhSQIgpCP5N32kaiQBEEQ8hVxDkkQBEHIFcQ5JEEQBCFXyLvVkaiQBEEQ8hXRZScIgiDkCqo83EYSFZIgCEI+IlpIgiAIQq4gBjXkEjY2RVi1ej5WVhYolUrWrfVm2bJ16vXDhvVjxsyxlCj+PVFR0k5Traeny4EjW9DT00Uul7NvzxFmzVgEwNgJI2jdtjlKhZK1q7ew0muDpNlvnA04QkJ8IgqFAkW6gpaNujBi9CC69mivPt7ZUxdx0veMpLmlypZg3srp6v8XK2nDktkrsa35LaXLlgTApIAxcbHxtG/UQ9LsN/ueuXACFSqVRaUCj58n8/eVGwD8NLgHv08eQc0KDYmJfil5dqmyJZi/aob6/8VLFmWx50o2rNwKQG/3boyaNAy7io15Gf1K8nyZTMYRvx2EhYbTs4s7Dk51mTD1N7RkMhLjExjmPpZHDx9Lltd8/BqM9HWQacmQa2uxZXQ39br1vldYsPsMJz0HUsjYgIdh0UzcdJSgJxEMcbOnl0tNScqgp6fLYZ+t6OrpIpdrs3fPEWZO/4N+A3rg7t6bMmVLUrpkTaIl/oy/7cL1o8THJ6BUKElPT8e1YWdMTQuyfO1cipew4cnjZwzs/SuvXsVqrAwfk3ero3xWISkU6Yz5fRoBAbcwNjbC/+x+Tpw4w507IdjYFKFhQ0ceP36qkeyUlFTatOxJQkIicrmcw0e34nvsNBW+KYuNTRHqVG+KSqWicGHN3gWgc6s+//niXe21kZVL1mss89H9x+qKRiaTcfL6AXwP+bHx9ZcywMhJPxMfm6CR/AkzRnL6xDmG9BmFjo4cfYOMq+6LFLWiXv26PHvyXCO5kHHs7Rp2BzKO3e/GQXwP+QFgXdQS+/p1CNVgfr9BPbgXfB8TE2MAZs2fQO8fhnDv7gN69e3C8JEDGO4+VtLMVcM6UsjY4J1lYTFxXLjzmCKFTNTLChrpM6pjA05evy9pfkpKKm4tuqs/az7HtnHs6Ckunr+Kz+ETHDi8RdK8j+no1vudz9rgET/hf/oiSxeuZvDwnxg84idmTJr/VcrytrzcQpJpOuDQoUNERkZqOgaAsLAXBATcAiA+PoHg4PsULWoNgOfs8YwbNxOVBt+rhIREAHR05Mh15KhUKnr37coczyWoXgdHRkZrrgC5QF3HWjx59JTnT8PeWd60lQsHdx+VPM/Y2IhadtXZvmkPAGlp6cTFxgMwdtqveE5eqH7tNa2uU8axh74+do+pI5g7ZbHG8osUtaJRk/ps2fiXeplKpcL4deVUoIAJ4c9faCT7fXN3+jG8jSNo/TsZj5mJId+WtEauLf3XzNufNZ3Xn7UbN27z+PEzybOyqmlzZ3Z4Z/wd7vDeQzPXhjlSDmU2H7mBxltI27ZtY+/evVhZWVGvXj3s7OwoUKCApmMpUaIYtraVuXw5ANcWLjwPDScwMEijmTKZjJNn9lC6TAnWrNrM1SvXKV2mBG3btaCFW2OiIqPxGDWVB/f/0Ui+SqVi018rQAWb1+9gy/qdAPT6qSvtO7fiRsAtpo2bq9FuhOZtG3PovYqnRt1qRL2I5vHDJ5LnFS9lQ3RUDLMXT6JilQrcvBHE1DFzsHeqTfjzCO7cuid55se4tmnMwV0Zx+7c1JHw5y8I1mD+lJkeTJswFyMTI/Wy336ewKYdXiQnJRMfl0CLxl0kzdTSgkFLdqEFtHf4jg4OVfG7cR8LU2O+KWYhadanyGQyTvnvpUyZkqxeuYmrV65/tWzI+Kx571qV8Zn7cweb1++gsKU5EeEZP74jwiMxt8iZeyLm5VF2Gm8hWVlZsXz5ctq3b8+DBw8YMWIE06dPx8/Pj6QkzdzA08jIkC3eyxk1agrp6emMGjWEqVM133RWKpXUr9eKbys6Ur1GVSpVKo+uri4pKSk0qt+ODeu3s3jZTI3lt2/ekxbOnenZaRA9+3ahtl0NNq7djmN1V5o5dSAi7AXjpv2msXwdHTnOTRzx2X/ineWubZv8p5KSilyuTZWqFdm8bietGv5AUkISw0YNxH1EXxbM8tJI5ofo6Mhp2NQJn/3H0TfQY8Dw3iz2XKGxPJem9Yl8Ec2N67ffWd7fvSfdOw6kRpWGbN28m0nTR0ua++cvndnq0Y2lg9uy/fR1rt57yuojl3BvaS9pTmaUSiWO9m5U/qYe1WvaUqlyha+a36ZZd5o16Ej3jgP58aeu1LGv8VXzP0W0kD5BS0sLmUyGra0ttra2pKenExAQgL+/Pxs3bmTNmjUffJ6vry++vr6fnSeXy9myxYttW/ewb68PVap8Q6mSxbhw8TAANjbWnD13gPpObQgP10x3RuyrOM6euUijxk6Ehoaxb68PAAf2HWXJslkayQQID8s4nqjIaHwOHqdajW+5dP6qer33hr9Yt3WJxvIdGtlzOzCYqBf/dktqa2vj0sKZTo17aSTzeWgEYaERXL92E4DD+48zbNQAipew4eCpjHNY1kUt2XdiM22b9CQyIkoj5XBsZM/twDtEvYimfKWyFCtRlD0nNwNgVdSSv3w30rlZb8nya9epTpPmzjRq4oSenh4mJkZs3LacchVK8/fVjAEd+3YfZsvOlZLkvWFpmtEdaGZiiLNtOa6GPOVZ1Cs6zdgEQMTLOLrO2symkV0pXNDoU7uSxKtXcfifuYCLixNBt+9qPO+Ntz9rhw/4Uq36d0RGRGFpVZiI8EgsrQq/8zn4mvJyC0njFdL7/edyuZyaNWtSs2ZNUlM/Pr2Ci4sLLi4uACxetPWj271v+XJPgoNDWLw4o6K7dSuYUqX+Hd1zO8gfRwc3yUfZmRc2Iy0tjdhXcejr61Hf2Z4/Fqzi0AFfnOrbsXnjTuo51CYk5KGkuW8YGBogk2mREJ+IgaEBjs72/DHHS/0BAWjashHBQSEayYcPt4TsnGrx8N4jwp9HaCQzMiKK58/CKV2uJA9D/sHeqTa3btyhR7uB6m1OXTtAG5fuGhll90aLtk3U3XX3gu7jUKWZep3vlT10aNJL0lF2M6YsYMaUBQDYOdRi0JDe9O42lBt3T1OmbEke3P8HJ2c77t2VbkBBUkoaSpUKI31dklLSOB/0DwOa1+Wk57+vdfPxa9gy+of/DHqQknlhM9LT0nj1+rPWwLkeC+drrjX6vvc/a/Ub2rNgthdHj5ykY9c2LF24mo5d2+Bz+ORXK9PbNNXaWbZsGdeuXaNgwYLMmzcPgO3bt3P8+HH1aZiuXbtSvXp1AHbv3s2JEyeQyWT07t2batWqZZqh8Qpp+PDhH12nq6sraZadXU1+6Naem4FBnL9wCIBJE2fj4+Mnac6HWFlZsGzFbLS1ZchkMvbsOszRIye5cP4KK9fMZ9DgH0lISGTYEGlHPL1hYWHOyo0LgYxurD07D3Hq+FkWLp9B5e8qolKpePr4Gb//MkUj+foGetg71Wbyb+92STZv899zSlKb/LsnC7ymo6Ojw5N/njJq6CSN5r1P30AP+/p1mPib5rpjs0KhUPDbsAms3vAHSpWSVy9jGTF4nGT7j4pL4JeV+wFIVyhpXqsi9aqU+uj2ka8S+GH2FhKSU9HS0mLzyb/ZNa4nxgZ6X1QOaysLvFbOQaatjUwmY/eug/gcOcmAQb0YNrwfVlYWnLtwkGM+fgwdMuaLsj7EwsKcNZsyLunQ1tZmz18H8Tvuz/VrgXitm0/X7u149vQ5A37MmfmllBoaRNOgQQOaNWvG0qVL31neokULWrVq9c6yp0+fcu7cOebPn09MTAxTp07ljz/+QCb79FkiLdXXGoL0BYwMS+VYtp62To5l5+SkYSY6hjmWDZCsSMuxbF1Zzl0N8TI1PseyH+z8Ocey4f97gr5nMbck21ePku2y9byN/+zKdJuIiAg8PT3faSHp6+v/p0LavXs3AG3btgVg+vTpdOzYkQoVPn2uL19dhyQIgvD/7mu3MHx8fDh9+jRlypShZ8+eGBsbEx0dTfny5dXbmJmZER2d+Tk1USEJgiDkI9m9MPb9gWRvn8f/mCZNmtChQwcg4xKfDRs24O7unu1r70SFJAiCkI9kd5RdViqg95mamqr/3ahRIzw9PQEwNzcnKurfEaXR0dGYmWV+XZbGr0MSBEEQvp6veR1STMy/o5UvXbpE8eLFAahZsybnzp0jLS2NiIgInj9/Trly5TLdn2ghCYIg5COaupfdwoULuX37NnFxcQwcOJBOnTpx69YtHj16hJaWFhYWFvTv3x+A4sWLY2dnxy+//IJMJqNv376ZjrADUSEJgiDkK5q6MPZDl/A0bPjx+/W1a9eOdu0+b8SfqJAEQRDykdxyG6DsEBWSIAhCPpIHLi39KFEhCYIg5CN5eT6kPFEhfVOwWI5lJyk+fr89TXsYG5b5RhoSJ9fMndizSl8u7W2lPkdkekqOZSuUOdfhUriV5m78mxUvH5/IfCMNKVKmWeYb5RGiy04QBEHIFcTdvgVBEIRcQXTZCYIgCLmCGNQgCIIg5AriHJIgCIKQK4hzSIIgCEKukJfPIYmbqwqCIAi5gmghCYIg5CN5eVBDvmshdevfiR1+G9l+cgMzlk1CV0+XWvWqs/noGraf3MDkP8aira2tkexSZUuw68Qm9ePy/RP07N9Fvb63ezeCIi5halZQ8uwVK+bw+PE1rl49pl5WtWplTp3aw8WLhzl79gA1a9pKngugp6eLr99fnDm/n3OXD+Mxdtg76z3nTuBJ2HWNZL8hk8k4evovNmxdBsCeQxs5dmYXx87s4u8gP9ZtXix5pp6eLif8duF//gAXLh/m99fH3W9AD/6+foJX8fcxMy8kee4bNjZFOHR4C1evHePyFR/c3X8EYP2GxZy7cJBzFw5yK+gM5y4c1FC2N1ev+XL5ylHc3Xu/s37YsH4kJD7CXMLjj42LZ8TYabh17YfbD/0JuBmEz4kztO42gO8cXLkZdFe9bVp6OmOmzqVtj0G4/dCfVRu2SVYOPT1djp7cid/ZffhfPMjoMRlTvzvWt+PE6d2c9N/LAR9vSpcpIVnm51CiytYjN8hXLSQL68J06duBDvW7k5KcyqwVU2jetjEDfuvDwE7DefzgCQNH9qVlp2bs9Zb+Q/ro/mPaNewOZHxB+t04iO8hPwCsi1piX78OoU+eS54LsHHjDpYvX8+aNQvUy2bMGMP06Qs5etSPpk2dmTFjDE2adJY8OyUlldYtepCQkIhcLufwsa34Hj3FlcsBVPv+WwoWLCB55vv6DerBveD7mJgYA9DGtYd63eoNC/E5JP1dAFJSUnFr0V193D7HtnHs6Ckunr+Kz+ETHDi8RfLMt6Ur0vn99+lcD7iFsbERZ87u58QJf3r1HKreZsbMscTGxkqerVCkM+b3aQS8zvY/u58TJ85w504INjZFaNjQkcePn0qaOWuhF/Xq1GTB9HGkpaWRlJyCibERC2eMZ/KcRe9se/TEGVLT0ti9cTlJycm07jYA18YNsCli9cXlSElJpW3Lnur3/eBRb3yPnWLugkl07+LOvbv36f3TD/wy0p2hgzy+OO9z5eVBDRpvIaWnp3Pq1Clu3LgBgL+/P2vWrOHIkSOkp6dLnqetrY2evh7a2toYGOiRlJhEWmoajx88AeDi6cs0atFA8tz31XWqxZNHTwl9mnH7H4+pI5g7ZbHGmtP+/peIiXn5zjKVSkWBAiYAFCxowvPn4RrJBkhISARAR0eOjo4OKpUKmUzGlOkeTBznqbFcgCJFrWjUpD5bNv71n3VGxobUc6rD4YPHNZL97nHLUalU3Lhxm8ePn2kk723hYS+4HnALgPj4BIKDQyhS1Pqdbdq1d2XH9v2SZ4eFvSDgnez7FH2d7Tl7POPGzUTKP/X4hASuXr9Je7emAOjo6FDAxJiypUpQuuR/by2mpaVFUnIy6ekKUlJS0dHRwdjIULLyvPO+yzPed5VKhUkBIwAKFDAh7HmEZHmfQ6lSZeuRG2i8hbRs2TIUCgUpKSmcOnWK5ORk6tSpQ2BgICEhIQwZMkSyrBdhkWz02sqhK3+RkpzC+VOXObrvBMPGu1PJ9huCrgfTqKUzVkUtJcv8GNc2jTm46ygAzk0dCX/+guBb9zSe+7bffpvMgQMbmTVrLFpaMpyd22osSyaT4ee/h9JlSrJm5SauXrnOAPdeHD54nPDwFxrLBZgy04NpE+ZiZGL0n3WuLV3wP3WB+LgEjWTLZDJO+e+lTJmSrH593DmhRAkbbG0rc+VygHpZvXq1iYiI5P79RxrOLoatbWUuXw7AtYULz0PDCQwMkjTj6bMwCpkWZNz0+QSHPKDyN+XxGD4QQwP9D27f2NmBE2fO49z6B5KTUxj1c38Kvv5xJgWZTMbx07spXaYEa1dt5tqVGwwfMo6tO1eRnJRCXFw8TRt1lCzvc+SOqiV7NN5Cevz4MSNGjGDkyJHcuHGDX3/9FScnJ9zd3Xn06JGkWSYFTWjQ1IGWdTrRtFobDAz1cW3fhN8HTuS3yT+z4dBKEuMTUaQrJM19n46OnIZNnfDZfxx9Az0GDO/NYs8VGs38kP79ezBy5BTKlavLqFFT8PKao7EspVKJk30rqnzjQPWattjXq0WbNs1Z6bVBY5kALk3rE/kimhvXb39wfZv2Ldjz1yGN5SuVShzt3aj8TT2q17SlUuUKGsv6GCMjQzZ7L2f0qKnExcWrl3fs5KaR1tH72Vu8lzNq1BTS09MZNWoIU6fOlzwnXaEg6G4Indu2YOefSzEw0GfNxu0f3T7wdjDaMhkn9m7myM4/We+9iyfPpOsuVyqVODu0pmolJ6rXqErFSuUZOPhHunToR9VKTnhv+otpM8ZIlvdZZcvD55A0XiGpVCrS09NJSkoiJSWFxMSMpm5aWhoKxccrBl9fXzw8PPDwyHofbB3Hmjx7/JyXUS9JT1dw4tBpqtb8jhtXb9G3zWB6uvbn2oUAHj+Utm/7fY6N7LkdeIeoF9EUL1WMYiWKsufkZnyv7MGqqCV/+W6ksKW5RssA0L17e/bsOQzAX39pblDD22JfxeF/5iIOTnUpXbYk124c5/otPwwNDbh6Xfpus9p1qtOkuTOXbhzDa808HJzqsGRFRhdhoUIFqVbjO3x9Tkme+75Xr+LwP3MBFxcnjWe9TS6Xs3nLcrZt3cu+vT7q5dra2rRq1Yy//jqg0ewtW7zYtnUP+/b6UKZMSUqVLMaFi4e5HeSPjY01Z88dwMrK4ouzrC0LY2VRmKpVKgLQpIEDt++GfHT7Q8f8qFe3JjpyOeaFTKlWtTK37kjfQxH7Ko6z/pdwaexEle8qcu1KxqmJ3bsOUavO95LnZUVerpA03mXn7OzM8OHDUSqVdOnShfnz52Npacm9e/ewt7f/6PNcXFxwcXEB4Oh6hyxlhT0L57saVdA30CM5KYXaDjW4ff0OhcxNiYl6iY6uDj8O7saaPzT7q71F2ybq7rp7QfdxqPLvre19r+yhQ5NevIx+pdEyADx/Ho6TU11On76As3M9QkIeaSTHvLAZaWlpxL6KQ19fjwbO9vwxfyUVy9qpt3kSdp0ato0kz54xZQEzpmQM5LBzqMWgIb0ZMmA0AG5tmuHr40dKimamEDEvbEZ6Whqv1Mddj4Xzv25LeNlyT4KDQ1iyeM07y50b1uPu3fuEPtPcFCbLX2cvfp1961YwpUrVVK+/HeSPo4MbUVExX5xV2NwMa0sLHv7zlNIli3HhagBlS+UUeXcAABrCSURBVH18FFsRKwsuXb2OW9OGJCWncOPWHXp0kqbL2ty8EGnp6eq/d6cG9ixeuJICBUwoW64U90Me0cC5HneD70uS97ny8rBvjVdILVu2VFc8ZmZm1K9fn8DAQFxcXChXrpykWTf/vs3xAyfZfHQtinQFwTfvsmvTPgaP7odjY3u0tGTs3LCby2evSZr7Nn0DPezr12HibzM1lvEhGzYsxtHRjsKFCxEScpFp0+bj7u7B3LmTkMu1SU5OYfBgzYz4sbayYNnKOWhry5DJZOzedQifIyc1kvU5WrdvzpIFqzW2f2srC7xWzkGmrf36uA/ic+QkAwb1YtjwflhZWXDuwkGO+fgxdIj03Td2djX5oVs7bgbeUQ/tnjRxDkd9/OjQwY0dO/ZJnvludntuBgZx/sKh19mz8fHx01jmmBGDGD15NmnpaRQvWoSpY0bge+osMxcsJ/rlK9xHTqRi+TKsXDCdru3cGDdjPm26D0SFijauTfimXGlJymFlbckSL0/13/ve3Yc5esSPEUPHsW7jYpRKFa9evuLnwTnXZZdXaanyQHVavUjWWkia8P86QZ9BDk6QBzk7QV/i/+kEfTk9XPj/eYK+yNi7mW+URbWKZq/b+HLoacnKkF356jokQRCE/3d5oI3xUaJCEgRByEfycpedqJAEQRDyEdFCEgRBEHIF0UISBEEQcoWcHpzyJUSFJAiCkI/klvvSZUe+m35CEARByJtEC0kQBCEfEV12giAIQq6Ql7vs8kSF9Chec/P4ZGaAea0cy14Yl3PHXUBPurljsuNZXFSOZcu0tHIsu4ixWY5lJ6fn3F1JAOpV7Z35RhryeFL9HMuWmmghCYIgCLmCaCEJgiAIuYJoIQmCIAi5gmghCYIgCLmCaCEJgiAIuYJKlXNTmHwpUSEJgiDkI+JedoIgCEKuIO72LQiCIOQKooWUS+jp6XLgyBb09HSRy+Xs23OEWTMWscTLk3r1ahEbGw/A4IGjuRkYJElm+9n9qdjwe+KjYvmj6WgArCuVoO30vuga6hHzNJJtw5eSEp+Usa5icdrO+Ak9YwNUSiVLW48nPSXti8tRrFgRVq9egJWVBUqlkrVrt7B06TpmzBiDq2sjUlPTePjwH/r3H8mrV7FfnPc+/78PkxCfiEKhQKFQ4NaoK5W//Ybp88ajp6eLQqFg3MjpXL92U/LsVSvn0cLVhYgXkVT7vhEAhQqZ/q+9O4+P6d7/OP6amSRCSCSTBRFbhApXUCpBYonQhohLm1Lql6LWNFe1CNqKrU2QlrjxSxdLY6lL3etaS2OpUiXNcu2RhBIislkSWSYzc35/5Gcqpa7rzpyM+D77yOMh52TO+zunnM98v+eb8+Wbjf9L8+ZuXLmSzYg3JnH79h2jZ3/x+TICA/uTn19A5y79q+17992JREd9SOMmf6Kw8JbRs+HR5x0g9O2RjBk/Ep1Wy8H9P/LJ/M9Mkq9UKtn/w7fk5uQx+vVJrPpyKV6dO6CtrCQ1+TTvT5uHVqs1eu7r44YzdNRgFAoF2zfuYvNX3xr2jZr0On/5aAoBHYZwp8iI/88VCqxHfYBUcpuK7SupEzIThZV11b56DdDnXkazYxUqdy8sewwFSULS66g8/Df0OZnGa8e/8Sz3kGrVw1UrKjQMHTwGvx5D8OsxBP/+fnTt1gmAeR8soXfPIfTuOcRoxQgg+dsjrP2f6Grbhke9zXfR37Di5QjO7kvCb8JgAJQqJSGfTeUfc1ezfMBMvhyxCF2lcf6xarU6IiIW0bmzP717D2XixDG88IIHBw78yIsvDuCll14mI+MyM2ZMMUreo4wIHkdgnxDDRXF25LusWBJPYJ8QPv0kjtnz3jVJbkLCFgYNHlVt26yZUzl46Cjt2vfi4KGjzJo51TTZ67cyOGj0Q9ubNm2Mv78vV65cM0nug35/3n16dSPglb687DucgJ7D+CLua5Nlvz15DBnplwzfb9uyk55dX6G3zxCs61oz6n9eNXpmq7YtGTpqMKGDJjGq/zh6Bfjg1tIVAOcmTnT368qNa7lGz7Xo3B990Q3D9xVbllC+YQHlGxagv5GFLiMVAN3VC5Svn0/5hgVo9q/DasAYo7flcfSS9FRf5qBWFSSAe/dKAbC0tMDC0sLknxZ+PXmB0jsl1bY5tmrM5RMXAMg8epr2r1Q9fsjDtyO5F66Se/4qAKW3S5D0xmlfbm4eaWlVvY+SkntcuJBJkyYuHDjwIzqdDoCTJ1NxdW1slLwnIUkS9RvYANDAtgF5ufkmyfnx6AmKbt2uti0oaCAJ67cCVUVjyJCXTZJ99OgJbv0uG2DZ0kjmzF5cI59WR78VwqoVq9FoqnrehQVFJslp3MSFgIG92Ziw1bDtwPdHDH9OTT5FkyaNjJ7b0qM5Z1LOUVFWgU6nI+X4v+jzih8A70aGsXJRvNHPu6K+PapWf0J7+ujDOy3roHJ7AV1WVUGisuK311nWQe4RNOkp/zMHshSk3NxcduzYwdq1a0lISGD//v2UlpaaJEupVPLDsR2kX/qZw4eOkfzLvwCYO+9dfjy+k8WfzMHKysok2ffdvHiNdgEvAvCnQG8aNlYD4NiqEUgSbyVEELZrMX4TB5skv1mzpnTq1J6kpLRq28eMCWHfvsMmyUSCDd9+zq4Dmxk5ZjgAC+YuYc786Rw/tZ+5C6YTvXCFabIfwcXZkdzcPKCqWDs7qWXLHjw4gOs5uZwyYk/8Dz3ivLd0b85L3i+yff9G/rZjDR07tzdJ9MKoOSz4aBn6R3yosrCw4NURQziY+KPRc7MuXKZzdy/s7G2pU7cOPft549LEGd8BPcjPLSDjXJbRMy37vI7myLfwiCnVqtZd0F29AJryB7Z1xjp0AXX+HI5m/zqjt+dxJEl6qi9zYPJ7SHv27CE5ORlPT0+ysrJo0aIFhYWFzJ07l/Hjx9O+vXH/sej1enr3HIKtXQPWb1pFu3YeLJy3jJs387GysuKzlQv5y7sTWBr9V6PmPmjbzC8ImjcG//BhnE9MNgzLKVUqmndrS9yQD6ksq2D8prlcP32ZrJ/OGi3bxqYe33wTz4wZCygu/q3nNnNmGDqdls2b/2G0rAcNCxxDXm4+akcHNmz7nKyMXwkc0p+FHyxl785EBgUPYEnsfEYNm2CSfHNRt641EbPCCRz0hix5jzrvFhYW2DVswNABo/Dq0oFVq5fRq8srRs0NGNiHgvxCTqWdpUevlx7aH/3pR/x87BdOHE82ai7Ar5lXSFi1iZWbYyi7V0bGuUx0Wi1vhb/JOyPfN3qesmVHpNK7SHlXUTRt89B+ixe6oT1Tveeky0xFl5mK0tUDyx7BVGwzzT28R3mWJzWYvId04MAB5s6dy/Dhw/nwww+5du0aI0eOZO7cuXz99R+PbScmJhIREUFERMRT5d69U8yxH0/gH+DHzZtVQ0UajYZN67fRpWvHpzrmk8rPymHNmCj+GjSXf+34icIrVZ/U7+QWcfnEeUpvFVNZriH9UBpNOrQ0Wq6FhQXffBPP3/62nX/+8zvD9lGjhhMY6E9o6F+MlvV794fjCguK2Lf7IJ26dGD4iCHs3ZkIwO5/7serSweT5f/ezbwCGjVyBqBRI2fy8uV5erh7qxa0aOHGL0n7uZh+nKZNG3Pi5+9wcXEySd6jzvuNnJt8t+sAAP9KOYNer8dBbW/U3Je8uzDwlX4knTrA52ti6OnXnbgvlgDw3qypqNUOfDQnyqiZD9rxzR7GDHybicPCuXO7mJzsXJo0a8zGxNVsP7EZ58ZOrN/3JWqn//7p6SpXd1TunbAe9wl1Bk1A6dYWq1fGVe20tkHZqCW6S6ce+Vr99QwUDZ3Buv5/3Y4n9Sz3kGQZsrt/D6OyspKysqrZZo6Ojobtj9K/f3+ioqKIinryv9RqRwds7RoAYG1dh959e3Dx4qVqF4NBgwM4f+7i07yNJ2ajtgVAoVDQN+zPnNhYdVG++MMpGr3QDEtrK5QqJS27tyMvw3g3vePjl5Cenkls7FeGbQEBvXnvvcm8+uo4ysrKH/Pqp1e3Xl1s6tcz/Nmvrw/p5zPJy83Hu2dXAHr6defXrKsmyX+UXTv3M+bN1wAY8+Zr7Ny5T5bcM2cv0NStE23a+tCmrQ/Xrt2gu/fLhg9FxvRH533/noP08K3qtbR0b46llSVFRp7lt3j+p3T27EO3jv5MHPsex46cYOqEmYwa8yp9/Xsxadx7Jr3I2asbAuDi6kzfQF/2fLuPlzsOZWj3EQztPoK8G/m8OfBtCvP/+/tnlUf/QfmXMylfPZuK3V+gz05Hs3c1ABZtulYVI91vk5MUDX+73iicm4FKBeUlDx1XeJjJh+z8/f2ZPXs2Hh4enD9/nuDgYADu3r1L/frG/dTg4uLEqs+XoFIpUSqVbP/7XvZ/d4jtuxJwdHRAoVBw+tR53pv2kdEyR8SG0dK7HTb2DYg4vpLEz7ZhZWONz5sBAJzZl0Ty1h8AKL97j6Nf7WHqjkVIkkT6oTTSD6U97vBPrEeProwaNZzTp8/z8897AJg3bykxMZFV0+F3bQCqJjaEh881SuZ9jk4OfJGwHAALCxX/3LaXHw4eY9a0UiI/noXKQkVFhYaI6fONmnvfhvVx9PbzwdHRgV8v/cL8BcuIXhrH5k3xvBU6kuzs67w+cqJJstcn/BW//8++lJXEgoUxrFu32SRZv/dH593S0oKlKxew/+jfqdRU8t7UD2RpD8CSzyK5lp3D7u+rzsHund/z6ZJVRs+J/mohtva26Cq1LJ2znOI7NXPBV7XtRmXS3urbPF7Eop0P6HVIWg2aXV/I2iZzmTH3NBSSDH217Oxsrl+/jpubG66urv/x6x0aeJigVU+mRhfou3msxrKd69nVWDaIBfpqQk0v0NfcxqXGsg+///C9ITnVm/6l0Y5lX7/1U73uVol8vyv1R2T5xVg3Nzfc3NzkiBIEQXiuPcuTGmrVkxoEQRCed+YyQeFpiIIkCIJQizzL95BEQRIEQahFzOWpC09DFCRBEIRaRPSQBEEQBLMg7iEJgiAIZkEM2QmCIAhmQfSQBEEQBLMgCpIgCIJgFp7dcgRIz4Hvv/9eZD9n+SL7+ct/XrNrk1q3YuyjJCYmiuznLF9kP3/5z2t2bfJcFCRBEATB/ImCJAiCIJgFVWRkZGRNN0IOrVq1EtnPWb7Ifv7yn9fs2kKW9ZAEQRAE4d8RQ3aCIAiCWajVv4eUlpbG2rVr0ev1+Pv7M3ToUNmyV61aRUpKCnZ2dsTExMiWC1BQUEBcXBy3b99GoVDQv39/AgMDZcnWaDTMmzcPrVaLTqfD29ubkJAQWbLv0+v1RERE4ODgQEREhKzZU6dOxdraGqVSiUqlIioqSrbse/fuER8fT3Z2NgqFgsmTJ9OmjelXQs3JyeGzzz4zfJ+Xl0dISAiDBg0yeTbArl27OHjwIAqFAjc3N6ZMmYKVlZUs2Xv27OHAgQNIkoS/v79s77nWquFp5yaj0+mksLAwKTc3V6qsrJTef/99KTs7W7b8s2fPSllZWdL06dNly7yvqKhIysrKkiRJkkpLS6Xw8HDZ3rter5fKysokSZKkyspKafbs2VJ6eros2fft3LlTWr58ufTJJ5/ImitJkjRlyhTpzp07sudKkiStXLlSSkxMlCSp6tyXlJTI3gadTieNHz9eysvLkyWvsLBQmjJlilRRUSFJkiTFxMRIhw4dkiX7ypUr0vTp06Xy8nJJq9VKCxYskHJycmTJrq1q7ZBdZmYmjRo1wsXFBQsLC3r06EFSUpJs+Z6entSvX1+2vAfZ29sbbrDWrVsXV1dXioqKZMlWKBRYW1sDoNPp0Ol0KBQKWbIBCgsLSUlJwd/fX7ZMc1BaWsr58+fp168fABYWFtjY2MjejtOnT9OoUSOcnJxky9Tr9Wg0GnQ6HRqNBnt7e1lyr1+/joeHB3Xq1EGlUtGuXTtOnjwpS3ZtVWuH7IqKilCr1Ybv1Wo1GRkZNdiimpGXl8fly5dp3bq1bJl6vZ5Zs2aRm5vLwIED8fDwkC173bp1jB49mrKyMtkyf2/x4sUABAQE0L9/f1ky8/LysLW1ZdWqVVy5coVWrVoRGhpq+HAgl2PHjtGzZ0/Z8hwcHAgKCmLy5MlYWVnh5eWFl5eXLNlubm5s3ryZ4uJirKysSE1Nxd3dXZbs2qrW9pCkR0welPOTujkoLy8nJiaG0NBQ6tWrJ1uuUqlk6dKlxMfHk5WVxdWrV2XJTU5Oxs7Orkan3y5cuJDo6GjmzJnDvn37OHfunCy5Op2Oy5cvM2DAAJYsWUKdOnXYvn27LNn3abVakpOT8fb2li2zpKSEpKQk4uLi+PzzzykvL+fIkSOyZDdt2pTg4GAWLVrExx9/TPPmzVEqa+0lVRa1toekVqspLCw0fF9YWChbV94caLVaYmJi8PX1pXv37jXSBhsbGzw9PUlLS6NZs2Ymz0tPT+eXX34hNTUVjUZDWVkZsbGxhIeHmzz7PgcHBwDs7Ozo1q0bmZmZeHp6mjxXrVajVqsNvVFvb2/ZC1JqaiotW7akYcOGsmWePn0aZ2dnbG1tAejevTsXL17Ez89Plvx+/foZhkk3bdpUbVRG+M/V2nLu7u7OjRs3yMvLQ6vV8tNPP9G1a9eabpYsJEkiPj4eV1dXBg8eLGv23bt3uXfvHlA14+706dO4urrKkv3GG28QHx9PXFwc06ZNo0OHDrIWo/LycsNQYXl5OadOnZKlEAM0bNgQtVpNTk4OUHWhbtq0qSzZ98k9XAfg6OhIRkYGFRUVSJIk6983gDt37gBVM1tPnjwp+/uvbWptD0mlUjF27FgWL16MXq+nb9++uLm5yZa/fPlyzp07R3FxMZMmTSIkJMTwScrU0tPTOXLkCM2aNWPGjBkAjBw5ki5dupg8+9atW8TFxaHX65EkCR8fH1588UWT55qDO3fusGzZMqBqCK1Xr1506tRJtvyxY8cSGxuLVqvF2dmZKVOmyJZdUVHBqVOnmDBhgmyZAB4eHnh7ezNr1ixUKhUtWrSQ7b4dQExMDMXFxVhYWDBu3Lgam8hUW4gnNQiCIAhmodYO2QmCIAjPFlGQBEEQBLMgCpIgCIJgFkRBEgRBEMyCKEiCIAiCWRAFSXhm3X+qtE6nA+Djjz/m8OHDJs/dsmULsbGxRj3m79+LXK8VBHNSa38PSTAPU6dO5fbt2yiVSqytrencuTNjx441yTPW5syZ88RtmjhxIh07djR6G86ePcvKlSuJj483+rEFobYTPSTB5GbNmsX69euJjo4mKyuLbdu2PfQzkiSh1+troHWCIJgL0UMSZOPg4ECnTp3Izs4GIDIykrZt23Lu3DkuXbpETEwMtra2fP3116SmpqJQKOjbty8hISEolUr0ej0bNmzghx9+oG7dug89FikyMhJfX1/D0hOJiYns3r2bwsJC1Go177zzDrt376agoIDo6GiUSiWvvvoqwcHBXLx4kYSEBK5du4aTkxOhoaG0b98eqBoSi4uL4/Lly3h4eNCkSZOnev8pKSls3ryZmzdvUq9ePcN7e9ChQ4fYunUrkiQRFBREUFAQUPUE9R07dnDgwAHu3btHhw4dmDBhgngygFCriIIkyKagoIDU1FReeuklw7YjR44wZ84cmjRpgiRJfPrppzRs2JDY2FgqKiqIiopCrVYTEBBAYmIiKSkpREdHY21t/diVeI8fP87WrVuZMWMG7u7u3Lx5E5VKxTvvvMOFCxeqDdkVFRURFRVFWFgYnTp14syZM8TExLB8+XJsbW1ZsWIFbdq04YMPPiAjI4OoqKinei5inTp1CAsLo2nTpmRnZ7No0SJatGhR7XycOXOGFStWkJeXx/z582nevDkdO3Zk7969JCUlERkZia2tLWvXruWrr75i2rRp/3E7BMFciSE7weSWLl1KaGgoH330EZ6engwbNsywr0+fPri5uaFSqSgpKSEtLc2wjo+dnR2DBg3ip59+AqqKTGBgII6OjtSvX/+xS9IfPHiQ4OBgWrdujUKheOyicUeOHKFz58506dIFpVJJx44dcXd3JyUlhYKCArKysnj99dextLTE09PzqZ/N1759e5o1a4ZSqaR58+b07NnzoeUpXnvtNaytrWnWrBl9+/bl2LFjQFVvb8SIEajVaiwtLXnttdc4ceKEmMgg1CqihySY3IwZM/5wAsGDj+svKChAp9NVe0CnJEmGn7l16xaOjo6GfY9blbSgoAAXF5cnal9BQQE///wzycnJhm06nY727dtTVFSEjY1NtUkYTk5OFBQUPNGxH5SRkcGmTZu4evUqWq0WrVb70NpBD54PR0dHw1pS+fn5LFu2rNqaXkql0vC0aUGoDURBEmrUgxdYtVqNhYUFq1evRqVSPfSz9vb21QrB44qCo6MjN2/efKI2qNVqfH19mTRp0kP78vPzuXfvHuXl5Yai9DTFCCA2NpaBAwcye/ZsrKysWLduHXfv3q32M4WFhYblEwoKCgxreKnVaiZPnswLL7zw0HHz8vKeqj2CYG7EkJ1gNuzt7fHy8iIhIYHS0lL0ej25ubmGYS0fHx/27t1LYWEhJSUlj12Arl+/fuzcuZNLly4hSRK5ubnk5+cDVWsHPXgR9/X1JTk5mbS0NPR6PRqNhrNnz1JYWIiTkxPu7u5s2bIFrVbLhQsXqvWk/ohGo6n2JUkSZWVl1K9fHysrKzIzMzl69OhDr9u2bRsVFRVkZ2dz+PBhevToAVQth75582bDe7h79y5JSUlPfnIF4RkgekiCWQkLC2Pjxo1Mnz6dsrIyXFxcCA4OBsDf35+cnBxmzJhB3bp1CQoK4syZM488jo+PD8XFxaxYsYKioiKcnZ0JCwvDycmJoUOHsmbNGjZs2MCwYcMYMmQIM2fOZMOGDaxYsQKlUknr1q15++23AQgPDycuLo633nqLNm3a4OfnZ1iE8FGKiooYPXp0tW2xsbGMHz+ehIQE1qxZg6enJz4+Pg8dx9PTk/DwcPR6PUFBQXh5eQEQGBgIwKJFi7h16xZ2dnb4+PjQrVu3pzvRgmCGxHpIgiAIglkQQ3aCIAiCWRAFSRAEQTALoiAJgiAIZkEUJEEQBMEsiIIkCIIgmAVRkARBEASzIAqSIAiCYBZEQRIEQRDMgihIgiAIgln4PwEUHJD3RQ+bAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 504x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "### BEST PCA\n",
    "trainer = Trainer(\n",
    "    units=1024,\n",
    "    activation_function=\"relu\", \n",
    "    optimizer=tf.keras.optimizers.Adam(learning_rate=2e-4), \n",
    "    n_layers=4,\n",
    "    loss_function = \"kl_divergence\",\n",
    "    input_size=200\n",
    ")\n",
    "trainer.train(x_train_processed_pca, y_train_processed, \n",
    "              batch_size=128, \n",
    "              epochs=500)\n",
    "trainer.print_summary()\n",
    "trainer.plot_history()\n",
    "trainer.evaluate(x_test_processed_pca, y_test_processed)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model_40\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " input_34 (InputLayer)       [(None, 1024)]            0         \n",
      "                                                                 \n",
      " dense_91 (Dense)            (None, 200)               205000    \n",
      "                                                                 \n",
      " dense_92 (Dense)            (None, 1024)              205824    \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 410,824\n",
      "Trainable params: 410,824\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n",
      "Epoch 1/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.2447 - val_loss: 0.1356\n",
      "Epoch 2/50\n",
      "313/313 [==============================] - 1s 5ms/step - loss: 0.1142 - val_loss: 0.1046\n",
      "Epoch 3/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0909 - val_loss: 0.0840\n",
      "Epoch 4/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0801 - val_loss: 0.0795\n",
      "Epoch 5/50\n",
      "313/313 [==============================] - 2s 6ms/step - loss: 0.0734 - val_loss: 0.0772\n",
      "Epoch 6/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0699 - val_loss: 0.0692\n",
      "Epoch 7/50\n",
      "313/313 [==============================] - 2s 6ms/step - loss: 0.0660 - val_loss: 0.0649\n",
      "Epoch 8/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0634 - val_loss: 0.0700\n",
      "Epoch 9/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0627 - val_loss: 0.0599\n",
      "Epoch 10/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0609 - val_loss: 0.0587\n",
      "Epoch 11/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0592 - val_loss: 0.0589\n",
      "Epoch 12/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0576 - val_loss: 0.0661\n",
      "Epoch 13/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0577 - val_loss: 0.0572\n",
      "Epoch 14/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0564 - val_loss: 0.0540\n",
      "Epoch 15/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0549 - val_loss: 0.0551\n",
      "Epoch 16/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0552 - val_loss: 0.0644\n",
      "Epoch 17/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0552 - val_loss: 0.0536\n",
      "Epoch 18/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0529 - val_loss: 0.0530\n",
      "Epoch 19/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0522 - val_loss: 0.0539\n",
      "Epoch 20/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0521 - val_loss: 0.0526\n",
      "Epoch 21/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0524 - val_loss: 0.0526\n",
      "Epoch 22/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0518 - val_loss: 0.0503\n",
      "Epoch 23/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0514 - val_loss: 0.0507\n",
      "Epoch 24/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0512 - val_loss: 0.0506\n",
      "Epoch 25/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0508 - val_loss: 0.0511\n",
      "Epoch 26/50\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0514 - val_loss: 0.0528\n",
      "Epoch 27/50\n",
      "304/313 [============================>.] - ETA: 0s - loss: 0.0497Restoring model weights from the end of the best epoch: 22.\n",
      "313/313 [==============================] - 2s 5ms/step - loss: 0.0497 - val_loss: 0.0537\n",
      "Epoch 27: early stopping\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "class TraineAutoencoder:\n",
    "    def __init__(self) -> None:\n",
    "        self.model = self.build_autoencoder_model()\n",
    "        self.history = None\n",
    "        self.training_time = None\n",
    "\n",
    "    def build_autoencoder_model(self):\n",
    "        self.input_ = tf.keras.Input(shape=(1024,))\n",
    "        self.encoded = tf.keras.layers.Dense(200, activation='tanh')(self.input_)\n",
    "        self.decoded = tf.keras.layers.Dense(1024)(self.encoded)\n",
    "        autoencoder = tf.keras.Model(self.input_, self.decoded)\n",
    "        autoencoder.compile(\n",
    "            optimizer=\"adam\", \n",
    "            loss='mean_squared_error'\n",
    "        )\n",
    "        return autoencoder\n",
    "    \n",
    "    def encode(self, x_test):\n",
    "        encoder = tf.keras.Model(self.input_, self.encoded)\n",
    "        encoded_inputs = encoder.predict(x_test)\n",
    "        return encoded_inputs\n",
    "    \n",
    "    def train(self, x_train, batch_size=128, epochs=50):\n",
    "        print(self.model.summary())\n",
    "        self.batch_size = batch_size\n",
    "        assert x_train.shape[1:] == (1024, )\n",
    "        early_stopping = tf.keras.callbacks.EarlyStopping(\n",
    "            monitor='val_loss', \n",
    "            verbose=1,\n",
    "            patience=5,\n",
    "            mode='min',\n",
    "            restore_best_weights=True\n",
    "        )\n",
    "        start = time()\n",
    "        self.history = self.model.fit(\n",
    "            x_train, x_train,\n",
    "            batch_size=batch_size,\n",
    "            epochs=epochs,\n",
    "            verbose=1,\n",
    "            validation_split=0.2,\n",
    "            callbacks=[early_stopping],\n",
    "        )\n",
    "        self.training_time = time() - start\n",
    "    \n",
    "    def plot_history(self):\n",
    "        fig = plt.figure(figsize=(12, 4))\n",
    "        metrics = ['loss']\n",
    "        for n, metric in enumerate(metrics):\n",
    "            plt.subplot(1, 1, n+1)\n",
    "            plt.plot(self.history.epoch, self.history.history[metric], label='Train')\n",
    "            plt.plot(self.history.epoch, self.history.history[f\"val_{metric}\"], linestyle=\"--\", label='Validation')\n",
    "            plt.xlabel('Epoch')\n",
    "            plt.ylabel(metric)\n",
    "            plt.title(metric)\n",
    "        plt.legend()\n",
    "        plt.show()\n",
    "        \n",
    "ta = TraineAutoencoder()\n",
    "ta.train(x_train_processed, batch_size=128, epochs=50)\n",
    "ta.plot_history()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(50000, 200)"
      ]
     },
     "execution_count": 169,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train_processed_autoencoder = ta.encode(x_train_processed)\n",
    "x_test_processed_autoencoder = ta.encode(x_test_processed)\n",
    "x_train_processed_autoencoder.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/500\n",
      "313/313 [==============================] - 4s 11ms/step - loss: 1.9939 - accuracy: 0.2854 - val_loss: 1.8223 - val_accuracy: 0.3554\n",
      "Epoch 2/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 1.7659 - accuracy: 0.3732 - val_loss: 1.7020 - val_accuracy: 0.3976\n",
      "Epoch 3/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 1.6254 - accuracy: 0.4227 - val_loss: 1.6144 - val_accuracy: 0.4304\n",
      "Epoch 4/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 1.5034 - accuracy: 0.4686 - val_loss: 1.5572 - val_accuracy: 0.4497\n",
      "Epoch 5/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 1.3956 - accuracy: 0.5059 - val_loss: 1.5346 - val_accuracy: 0.4569\n",
      "Epoch 6/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 1.3006 - accuracy: 0.5400 - val_loss: 1.5199 - val_accuracy: 0.4666\n",
      "Epoch 7/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 1.2135 - accuracy: 0.5701 - val_loss: 1.4963 - val_accuracy: 0.4735\n",
      "Epoch 8/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 1.1311 - accuracy: 0.5975 - val_loss: 1.4933 - val_accuracy: 0.4822\n",
      "Epoch 9/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 1.0561 - accuracy: 0.6251 - val_loss: 1.5313 - val_accuracy: 0.4746\n",
      "Epoch 10/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 0.9811 - accuracy: 0.6525 - val_loss: 1.5460 - val_accuracy: 0.4807\n",
      "Epoch 11/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 0.9099 - accuracy: 0.6788 - val_loss: 1.5723 - val_accuracy: 0.4811\n",
      "Epoch 12/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 0.8407 - accuracy: 0.7014 - val_loss: 1.5806 - val_accuracy: 0.4803\n",
      "Epoch 13/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 0.7723 - accuracy: 0.7286 - val_loss: 1.6265 - val_accuracy: 0.4856\n",
      "Epoch 14/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 0.7150 - accuracy: 0.7474 - val_loss: 1.6752 - val_accuracy: 0.4796\n",
      "Epoch 15/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 0.6582 - accuracy: 0.7697 - val_loss: 1.7338 - val_accuracy: 0.4799\n",
      "Epoch 16/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 0.6039 - accuracy: 0.7860 - val_loss: 1.7799 - val_accuracy: 0.4845\n",
      "Epoch 17/500\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 0.5611 - accuracy: 0.8013 - val_loss: 1.8382 - val_accuracy: 0.4771\n",
      "Epoch 18/500\n",
      "313/313 [==============================] - ETA: 0s - loss: 0.5095 - accuracy: 0.8211Restoring model weights from the end of the best epoch: 8.\n",
      "313/313 [==============================] - 3s 10ms/step - loss: 0.5095 - accuracy: 0.8211 - val_loss: 1.8749 - val_accuracy: 0.4766\n",
      "Epoch 18: early stopping\n",
      "######################\n",
      "Training Time: 56.96s\n",
      "batch size: 128\n",
      "n_layers: 4\n",
      "optimizer: <keras.optimizer_v2.adam.Adam object at 0x00000175BA7611D0>\n",
      "activation_function: relu\n",
      "loss_function kl_divergence\n",
      "Model: \"sequential_9\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " dense_93 (Dense)            (None, 1024)              205824    \n",
      "                                                                 \n",
      " dropout_36 (Dropout)        (None, 1024)              0         \n",
      "                                                                 \n",
      " dense_94 (Dense)            (None, 512)               524800    \n",
      "                                                                 \n",
      " dropout_37 (Dropout)        (None, 512)               0         \n",
      "                                                                 \n",
      " dense_95 (Dense)            (None, 256)               131328    \n",
      "                                                                 \n",
      " dropout_38 (Dropout)        (None, 256)               0         \n",
      "                                                                 \n",
      " dense_96 (Dense)            (None, 128)               32896     \n",
      "                                                                 \n",
      " dropout_39 (Dropout)        (None, 128)               0         \n",
      "                                                                 \n",
      " dense_97 (Dense)            (None, 10)                1290      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 896,138\n",
      "Trainable params: 896,138\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "313/313 [==============================] - 1s 3ms/step - loss: 1.4857 - accuracy: 0.4720\n",
      "Test Loss: 1.4856510162353516 Test Accuracy: 0.47200000286102295\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.59      0.46      0.51      1000\n",
      "           1       0.58      0.52      0.55      1000\n",
      "           2       0.40      0.30      0.34      1000\n",
      "           3       0.33      0.26      0.29      1000\n",
      "           4       0.37      0.44      0.40      1000\n",
      "           5       0.44      0.38      0.41      1000\n",
      "           6       0.47      0.56      0.51      1000\n",
      "           7       0.53      0.57      0.55      1000\n",
      "           8       0.51      0.67      0.58      1000\n",
      "           9       0.50      0.56      0.53      1000\n",
      "\n",
      "    accuracy                           0.47     10000\n",
      "   macro avg       0.47      0.47      0.47     10000\n",
      "weighted avg       0.47      0.47      0.47     10000\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "### BEST AC\n",
    "trainer = Trainer(\n",
    "    units=1024,\n",
    "    activation_function=\"relu\", \n",
    "    optimizer=tf.keras.optimizers.Adam(learning_rate=2e-4), \n",
    "    n_layers=4,\n",
    "    loss_function = \"kl_divergence\",\n",
    "    input_size=200\n",
    ")\n",
    "trainer.train(x_train_processed_autoencoder, y_train_processed, \n",
    "              batch_size=128, \n",
    "              epochs=500)\n",
    "trainer.print_summary()\n",
    "trainer.plot_history()\n",
    "trainer.evaluate(x_test_processed_autoencoder, y_test_processed)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## g)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## d)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>date</th>\n",
       "      <th>price</th>\n",
       "      <th>bedrooms</th>\n",
       "      <th>bathrooms</th>\n",
       "      <th>sqft_living</th>\n",
       "      <th>sqft_lot</th>\n",
       "      <th>floors</th>\n",
       "      <th>waterfront</th>\n",
       "      <th>view</th>\n",
       "      <th>condition</th>\n",
       "      <th>sqft_above</th>\n",
       "      <th>sqft_basement</th>\n",
       "      <th>yr_built</th>\n",
       "      <th>yr_renovated</th>\n",
       "      <th>street</th>\n",
       "      <th>city</th>\n",
       "      <th>statezip</th>\n",
       "      <th>country</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.833514</td>\n",
       "      <td>-0.423817</td>\n",
       "      <td>-0.441074</td>\n",
       "      <td>-0.843112</td>\n",
       "      <td>-0.829881</td>\n",
       "      <td>-0.193413</td>\n",
       "      <td>-0.022414</td>\n",
       "      <td>-0.084995</td>\n",
       "      <td>-0.309161</td>\n",
       "      <td>-0.667040</td>\n",
       "      <td>-0.565162</td>\n",
       "      <td>-0.672391</td>\n",
       "      <td>-0.530956</td>\n",
       "      <td>1.221538</td>\n",
       "      <td>-1.721431</td>\n",
       "      <td>-0.932755</td>\n",
       "      <td>-1.580132</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.833514</td>\n",
       "      <td>3.249245</td>\n",
       "      <td>1.759513</td>\n",
       "      <td>0.432754</td>\n",
       "      <td>1.568358</td>\n",
       "      <td>-0.161700</td>\n",
       "      <td>0.906456</td>\n",
       "      <td>-0.084995</td>\n",
       "      <td>4.829554</td>\n",
       "      <td>2.286168</td>\n",
       "      <td>1.789365</td>\n",
       "      <td>-0.069121</td>\n",
       "      <td>-1.674511</td>\n",
       "      <td>-0.825604</td>\n",
       "      <td>-1.720665</td>\n",
       "      <td>-0.823617</td>\n",
       "      <td>-1.529230</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-1.833514</td>\n",
       "      <td>-0.372384</td>\n",
       "      <td>-0.441074</td>\n",
       "      <td>-0.205179</td>\n",
       "      <td>-0.217344</td>\n",
       "      <td>-0.080969</td>\n",
       "      <td>-0.951284</td>\n",
       "      <td>-0.084995</td>\n",
       "      <td>-0.309161</td>\n",
       "      <td>0.809564</td>\n",
       "      <td>0.119158</td>\n",
       "      <td>-0.672391</td>\n",
       "      <td>-0.160982</td>\n",
       "      <td>-0.825604</td>\n",
       "      <td>-1.719899</td>\n",
       "      <td>-0.714478</td>\n",
       "      <td>-1.478328</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-1.833514</td>\n",
       "      <td>-0.234046</td>\n",
       "      <td>-0.441074</td>\n",
       "      <td>0.113788</td>\n",
       "      <td>-0.144670</td>\n",
       "      <td>-0.190125</td>\n",
       "      <td>-0.951284</td>\n",
       "      <td>-0.084995</td>\n",
       "      <td>-0.309161</td>\n",
       "      <td>0.809564</td>\n",
       "      <td>-0.959517</td>\n",
       "      <td>1.482145</td>\n",
       "      <td>-0.261884</td>\n",
       "      <td>-0.825604</td>\n",
       "      <td>-1.719134</td>\n",
       "      <td>-0.605340</td>\n",
       "      <td>-1.427425</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.833514</td>\n",
       "      <td>-0.003481</td>\n",
       "      <td>0.659220</td>\n",
       "      <td>0.432754</td>\n",
       "      <td>-0.206962</td>\n",
       "      <td>-0.121293</td>\n",
       "      <td>-0.951284</td>\n",
       "      <td>-0.084995</td>\n",
       "      <td>-0.309161</td>\n",
       "      <td>0.809564</td>\n",
       "      <td>-0.797135</td>\n",
       "      <td>1.051238</td>\n",
       "      <td>0.175357</td>\n",
       "      <td>1.208264</td>\n",
       "      <td>-1.718368</td>\n",
       "      <td>-0.496201</td>\n",
       "      <td>-1.376523</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4595</th>\n",
       "      <td>1.616831</td>\n",
       "      <td>-0.432390</td>\n",
       "      <td>-0.441074</td>\n",
       "      <td>-0.524145</td>\n",
       "      <td>-0.653387</td>\n",
       "      <td>-0.236663</td>\n",
       "      <td>-0.951284</td>\n",
       "      <td>-0.084995</td>\n",
       "      <td>-0.309161</td>\n",
       "      <td>0.809564</td>\n",
       "      <td>-0.367985</td>\n",
       "      <td>-0.672391</td>\n",
       "      <td>-0.564590</td>\n",
       "      <td>1.194991</td>\n",
       "      <td>1.739401</td>\n",
       "      <td>-0.823617</td>\n",
       "      <td>-1.580132</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4596</th>\n",
       "      <td>1.616831</td>\n",
       "      <td>-0.031267</td>\n",
       "      <td>-0.441074</td>\n",
       "      <td>0.432754</td>\n",
       "      <td>-0.705297</td>\n",
       "      <td>-0.202860</td>\n",
       "      <td>0.906456</td>\n",
       "      <td>-0.084995</td>\n",
       "      <td>-0.309161</td>\n",
       "      <td>-0.667040</td>\n",
       "      <td>-0.425978</td>\n",
       "      <td>-0.672391</td>\n",
       "      <td>0.410795</td>\n",
       "      <td>1.225622</td>\n",
       "      <td>1.740167</td>\n",
       "      <td>-0.605340</td>\n",
       "      <td>-0.969304</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4597</th>\n",
       "      <td>1.616831</td>\n",
       "      <td>-0.239536</td>\n",
       "      <td>-0.441074</td>\n",
       "      <td>0.432754</td>\n",
       "      <td>0.903911</td>\n",
       "      <td>-0.218438</td>\n",
       "      <td>0.906456</td>\n",
       "      <td>-0.084995</td>\n",
       "      <td>-0.309161</td>\n",
       "      <td>-0.667040</td>\n",
       "      <td>1.371813</td>\n",
       "      <td>-0.672391</td>\n",
       "      <td>1.285278</td>\n",
       "      <td>-0.825604</td>\n",
       "      <td>1.740932</td>\n",
       "      <td>1.031739</td>\n",
       "      <td>0.150549</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4598</th>\n",
       "      <td>1.667571</td>\n",
       "      <td>-0.618201</td>\n",
       "      <td>0.659220</td>\n",
       "      <td>-0.205179</td>\n",
       "      <td>-0.051232</td>\n",
       "      <td>-0.229139</td>\n",
       "      <td>-0.951284</td>\n",
       "      <td>-0.084995</td>\n",
       "      <td>-0.309161</td>\n",
       "      <td>-0.667040</td>\n",
       "      <td>-0.878326</td>\n",
       "      <td>1.525235</td>\n",
       "      <td>0.108089</td>\n",
       "      <td>-0.825604</td>\n",
       "      <td>1.741698</td>\n",
       "      <td>-0.823617</td>\n",
       "      <td>1.728524</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4599</th>\n",
       "      <td>1.667571</td>\n",
       "      <td>-0.587695</td>\n",
       "      <td>-0.441074</td>\n",
       "      <td>0.432754</td>\n",
       "      <td>-0.674151</td>\n",
       "      <td>-0.188118</td>\n",
       "      <td>0.906456</td>\n",
       "      <td>-0.084995</td>\n",
       "      <td>-0.309161</td>\n",
       "      <td>0.809564</td>\n",
       "      <td>-0.391183</td>\n",
       "      <td>-0.672391</td>\n",
       "      <td>0.646233</td>\n",
       "      <td>-0.825604</td>\n",
       "      <td>1.742464</td>\n",
       "      <td>1.577432</td>\n",
       "      <td>-1.478328</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4600 rows × 18 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          date     price  bedrooms  bathrooms  sqft_living  sqft_lot  \\\n",
       "0    -1.833514 -0.423817 -0.441074  -0.843112    -0.829881 -0.193413   \n",
       "1    -1.833514  3.249245  1.759513   0.432754     1.568358 -0.161700   \n",
       "2    -1.833514 -0.372384 -0.441074  -0.205179    -0.217344 -0.080969   \n",
       "3    -1.833514 -0.234046 -0.441074   0.113788    -0.144670 -0.190125   \n",
       "4    -1.833514 -0.003481  0.659220   0.432754    -0.206962 -0.121293   \n",
       "...        ...       ...       ...        ...          ...       ...   \n",
       "4595  1.616831 -0.432390 -0.441074  -0.524145    -0.653387 -0.236663   \n",
       "4596  1.616831 -0.031267 -0.441074   0.432754    -0.705297 -0.202860   \n",
       "4597  1.616831 -0.239536 -0.441074   0.432754     0.903911 -0.218438   \n",
       "4598  1.667571 -0.618201  0.659220  -0.205179    -0.051232 -0.229139   \n",
       "4599  1.667571 -0.587695 -0.441074   0.432754    -0.674151 -0.188118   \n",
       "\n",
       "        floors  waterfront      view  condition  sqft_above  sqft_basement  \\\n",
       "0    -0.022414   -0.084995 -0.309161  -0.667040   -0.565162      -0.672391   \n",
       "1     0.906456   -0.084995  4.829554   2.286168    1.789365      -0.069121   \n",
       "2    -0.951284   -0.084995 -0.309161   0.809564    0.119158      -0.672391   \n",
       "3    -0.951284   -0.084995 -0.309161   0.809564   -0.959517       1.482145   \n",
       "4    -0.951284   -0.084995 -0.309161   0.809564   -0.797135       1.051238   \n",
       "...        ...         ...       ...        ...         ...            ...   \n",
       "4595 -0.951284   -0.084995 -0.309161   0.809564   -0.367985      -0.672391   \n",
       "4596  0.906456   -0.084995 -0.309161  -0.667040   -0.425978      -0.672391   \n",
       "4597  0.906456   -0.084995 -0.309161  -0.667040    1.371813      -0.672391   \n",
       "4598 -0.951284   -0.084995 -0.309161  -0.667040   -0.878326       1.525235   \n",
       "4599  0.906456   -0.084995 -0.309161   0.809564   -0.391183      -0.672391   \n",
       "\n",
       "      yr_built  yr_renovated    street      city  statezip  country  \n",
       "0    -0.530956      1.221538 -1.721431 -0.932755 -1.580132      NaN  \n",
       "1    -1.674511     -0.825604 -1.720665 -0.823617 -1.529230      NaN  \n",
       "2    -0.160982     -0.825604 -1.719899 -0.714478 -1.478328      NaN  \n",
       "3    -0.261884     -0.825604 -1.719134 -0.605340 -1.427425      NaN  \n",
       "4     0.175357      1.208264 -1.718368 -0.496201 -1.376523      NaN  \n",
       "...        ...           ...       ...       ...       ...      ...  \n",
       "4595 -0.564590      1.194991  1.739401 -0.823617 -1.580132      NaN  \n",
       "4596  0.410795      1.225622  1.740167 -0.605340 -0.969304      NaN  \n",
       "4597  1.285278     -0.825604  1.740932  1.031739  0.150549      NaN  \n",
       "4598  0.108089     -0.825604  1.741698 -0.823617  1.728524      NaN  \n",
       "4599  0.646233     -0.825604  1.742464  1.577432 -1.478328      NaN  \n",
       "\n",
       "[4600 rows x 18 columns]"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('data.csv')\n",
    "CAT_COLS = [\"date\", \"street\", \"city\", \"statezip\", \"country\"]\n",
    "for cat_col in CAT_COLS:\n",
    "    df[cat_col] = pd.factorize(df[cat_col])[0]\n",
    "df = (df-df.mean()) / df.std()\n",
    "\n",
    "corr = df.corr()\n",
    "f, ax = plt.subplots(figsize=(14, 10))\n",
    "ax = sns.heatmap(corr, annot=True)\n",
    "bottom, top = ax.get_ylim()\n",
    "ax.set_ylim(bottom + 0.5, top - 0.5)\n",
    "plt.show()\n",
    "\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## v)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<BarContainer object of 16 artists>"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.datasets import load_diabetes\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.linear_model import LinearRegression\n",
    "\n",
    "features = df.drop([\"price\", \"country\"], axis=1)\n",
    "regressor = LinearRegression()\n",
    "regressor.fit(features, df[\"price\"])\n",
    "intercept = regressor.intercept_\n",
    "\n",
    "reg_importance = pd.DataFrame(zip(features.columns, np.abs(regressor.coef_)), columns=[\"feature\", \"coef\"])\n",
    "reg_importance = reg_importance.sort_values(by='coef', ascending=True)\n",
    "plt.barh(reg_importance[\"feature\"], reg_importance[\"coef\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<BarContainer object of 16 artists>"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "features = df.drop([\"price\", \"country\"], axis=1)\n",
    "regressor = DecisionTreeRegressor()\n",
    "regressor.fit(features, df[\"price\"])\n",
    "\n",
    "reg_importance = pd.DataFrame(zip(features.columns, regressor.feature_importances_), columns=[\"feature\", \"importance\"])\n",
    "reg_importance = reg_importance.sort_values(by='importance', ascending=True)\n",
    "plt.barh(reg_importance[\"feature\"], reg_importance[\"importance\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "collapsed_sections": [],
   "name": "NN_HW2_MohsenFayyaz.ipynb",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}