CV_ClassificationExample.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.model_selection import StratifiedKFold\n",
    "from sklearn.metrics import classification_report, accuracy_score, f1_score, precision_score, confusion_matrix\n",
    "\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras import layers\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense\n",
    "from tensorflow.keras.models import Model, load_model\n",
    "from tensorflow.keras.layers import Dropout\n",
    "from tensorflow.keras.layers import Flatten\n",
    "from tensorflow.keras.layers import Conv1D\n",
    "from tensorflow.keras.layers import MaxPooling1D\n",
    "from tensorflow.keras.utils import to_categorical"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# =====Read data======\n",
    "# Dataset link-> https://www.kaggle.com/sulianova/cardiovascular-disease-dataset\n",
    "rootPath='DS1/'\n",
    "dfDS = pd.read_csv(rootPath+\"cardio.csv\", delimiter=';')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "#=====Split Dataset=======\n",
    "X = dfDS.iloc[:, 1:len(dfDS.columns)-1].values\n",
    "Y = dfDS[\"cardio\"].values\n",
    "xTrain, xTest, yTrain, yTest = train_test_split(X, Y, test_size=0.20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "#=====Feature Scaling======\n",
    "columnsToScale = ['age', 'height', 'weight', 'ap_hi', 'ap_lo']\n",
    "scaler = MinMaxScaler(feature_range=(0,1))\n",
    "xTrain = scaler.fit_transform(xTrain)\n",
    "xTest = scaler.transform(xTest)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "verbose, epochs, batch_size = 1, 100, 512\n",
    "activationFunction='relu'\n",
    "\n",
    "def getModel():\n",
    "    model = Sequential()\n",
    "    model.add(Dense(100, input_dim = xTrain.shape[1], activation=activationFunction))\n",
    "    model.add(Dense(75, activation=activationFunction))\n",
    "    model.add(Dense(50, activation=activationFunction))\n",
    "    model.add(Dense(25, activation=activationFunction))\n",
    "    model.add(Dense(2, activation='softmax'))\n",
    "    model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])\n",
    "    return model\n",
    "\n",
    "model = getModel()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def showResults(test, pred):\n",
    "    target_names = ['positive', 'negative']\n",
    "    print(classification_report(test, pred, target_names=target_names))\n",
    "    accuracy = accuracy_score(test, pred)\n",
    "    precision=precision_score(test, pred, average='weighted')\n",
    "    f1Score=f1_score(test, pred, average='weighted') \n",
    "    print(\"Accuracy  : {}\".format(accuracy))\n",
    "    print(\"Precision : {}\".format(precision))\n",
    "    print(\"f1Score : {}\".format(f1Score))\n",
    "    cm=confusion_matrix(test, pred)\n",
    "    print(cm) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Results for fold 1\n",
      "Train on 44800 samples, validate on 11200 samples\n",
      "Epoch 1/100\n",
      "44800/44800 [==============================] - 1s 13us/sample - loss: 0.6612 - accuracy: 0.6002 - val_loss: 0.6409 - val_accuracy: 0.6364\n",
      "Epoch 2/100\n",
      "44800/44800 [==============================] - 0s 6us/sample - loss: 0.6391 - accuracy: 0.6332 - val_loss: 0.6315 - val_accuracy: 0.6472\n",
      "Epoch 3/100\n",
      "44800/44800 [==============================] - 0s 6us/sample - loss: 0.6329 - accuracy: 0.6403 - val_loss: 0.6276 - val_accuracy: 0.6535\n",
      "Epoch 4/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6326 - accuracy: 0.6411 - val_loss: 0.6301 - val_accuracy: 0.6449\n",
      "Epoch 5/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6304 - accuracy: 0.6446 - val_loss: 0.6295 - val_accuracy: 0.6464\n",
      "Epoch 6/100\n",
      "44800/44800 [==============================] - 0s 6us/sample - loss: 0.6311 - accuracy: 0.6401 - val_loss: 0.6312 - val_accuracy: 0.6393\n",
      "Epoch 7/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6292 - accuracy: 0.6463 - val_loss: 0.6242 - val_accuracy: 0.6538\n",
      "Epoch 8/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6276 - accuracy: 0.6459 - val_loss: 0.6231 - val_accuracy: 0.6538\n",
      "Epoch 9/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6280 - accuracy: 0.6459 - val_loss: 0.6236 - val_accuracy: 0.6548\n",
      "Epoch 10/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6259 - accuracy: 0.6478 - val_loss: 0.6228 - val_accuracy: 0.6556\n",
      "Epoch 11/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6263 - accuracy: 0.6482 - val_loss: 0.6226 - val_accuracy: 0.6522\n",
      "Epoch 12/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6259 - accuracy: 0.6487 - val_loss: 0.6220 - val_accuracy: 0.6546\n",
      "Epoch 13/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6248 - accuracy: 0.6488 - val_loss: 0.6238 - val_accuracy: 0.6533\n",
      "Epoch 14/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6242 - accuracy: 0.6497 - val_loss: 0.6225 - val_accuracy: 0.6564\n",
      "Epoch 15/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6242 - accuracy: 0.6492 - val_loss: 0.6207 - val_accuracy: 0.6583\n",
      "Epoch 16/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6271 - accuracy: 0.6455 - val_loss: 0.6237 - val_accuracy: 0.6496\n",
      "Epoch 17/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6233 - accuracy: 0.6505 - val_loss: 0.6213 - val_accuracy: 0.6583\n",
      "Epoch 18/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6232 - accuracy: 0.6508 - val_loss: 0.6219 - val_accuracy: 0.6503\n",
      "Epoch 19/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6233 - accuracy: 0.6506 - val_loss: 0.6194 - val_accuracy: 0.6578\n",
      "Epoch 20/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6229 - accuracy: 0.6500 - val_loss: 0.6266 - val_accuracy: 0.6421\n",
      "Epoch 21/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6216 - accuracy: 0.6522 - val_loss: 0.6187 - val_accuracy: 0.6571\n",
      "Epoch 22/100\n",
      "44800/44800 [==============================] - 0s 4us/sample - loss: 0.6210 - accuracy: 0.6537 - val_loss: 0.6185 - val_accuracy: 0.6583\n",
      "Epoch 23/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6205 - accuracy: 0.6535 - val_loss: 0.6259 - val_accuracy: 0.6480\n",
      "Epoch 24/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6206 - accuracy: 0.6525 - val_loss: 0.6245 - val_accuracy: 0.6448\n",
      "Epoch 25/100\n",
      "44800/44800 [==============================] - 0s 6us/sample - loss: 0.6210 - accuracy: 0.6529 - val_loss: 0.6173 - val_accuracy: 0.6606\n",
      "Epoch 26/100\n",
      "44800/44800 [==============================] - 0s 6us/sample - loss: 0.6225 - accuracy: 0.6494 - val_loss: 0.6169 - val_accuracy: 0.6615\n",
      "Epoch 27/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6187 - accuracy: 0.6558 - val_loss: 0.6186 - val_accuracy: 0.6556\n",
      "Epoch 28/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6203 - accuracy: 0.6527 - val_loss: 0.6164 - val_accuracy: 0.6601\n",
      "Epoch 29/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6204 - accuracy: 0.6513 - val_loss: 0.6179 - val_accuracy: 0.6546\n",
      "Epoch 30/100\n",
      "44800/44800 [==============================] - 0s 6us/sample - loss: 0.6195 - accuracy: 0.6535 - val_loss: 0.6179 - val_accuracy: 0.6547\n",
      "Epoch 31/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6174 - accuracy: 0.6559 - val_loss: 0.6175 - val_accuracy: 0.6581\n",
      "Epoch 32/100\n",
      "44800/44800 [==============================] - 0s 4us/sample - loss: 0.6174 - accuracy: 0.6546 - val_loss: 0.6147 - val_accuracy: 0.6617\n",
      "Epoch 33/100\n",
      "44800/44800 [==============================] - 0s 4us/sample - loss: 0.6165 - accuracy: 0.6573 - val_loss: 0.6151 - val_accuracy: 0.6608\n",
      "Epoch 34/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6155 - accuracy: 0.6571 - val_loss: 0.6194 - val_accuracy: 0.6525\n",
      "Epoch 35/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6150 - accuracy: 0.6586 - val_loss: 0.6142 - val_accuracy: 0.6586\n",
      "Epoch 36/100\n",
      "44800/44800 [==============================] - 0s 4us/sample - loss: 0.6141 - accuracy: 0.6614 - val_loss: 0.6132 - val_accuracy: 0.6581\n",
      "Epoch 37/100\n",
      "44800/44800 [==============================] - 0s 4us/sample - loss: 0.6198 - accuracy: 0.6531 - val_loss: 0.6118 - val_accuracy: 0.6667\n",
      "Epoch 38/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6138 - accuracy: 0.6600 - val_loss: 0.6125 - val_accuracy: 0.6629\n",
      "Epoch 39/100\n",
      "44800/44800 [==============================] - 0s 7us/sample - loss: 0.6131 - accuracy: 0.6580 - val_loss: 0.6119 - val_accuracy: 0.6669\n",
      "Epoch 40/100\n",
      "44800/44800 [==============================] - 0s 4us/sample - loss: 0.6127 - accuracy: 0.6608 - val_loss: 0.6136 - val_accuracy: 0.6644\n",
      "Epoch 41/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6114 - accuracy: 0.6623 - val_loss: 0.6099 - val_accuracy: 0.6642\n",
      "Epoch 42/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6088 - accuracy: 0.6650 - val_loss: 0.6105 - val_accuracy: 0.6643\n",
      "Epoch 43/100\n",
      "44800/44800 [==============================] - 0s 6us/sample - loss: 0.6065 - accuracy: 0.6693 - val_loss: 0.6046 - val_accuracy: 0.6735\n",
      "Epoch 44/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6083 - accuracy: 0.6673 - val_loss: 0.6038 - val_accuracy: 0.6732\n",
      "Epoch 45/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6040 - accuracy: 0.6716 - val_loss: 0.6004 - val_accuracy: 0.6764\n",
      "Epoch 46/100\n",
      "44800/44800 [==============================] - 0s 4us/sample - loss: 0.6020 - accuracy: 0.6746 - val_loss: 0.6191 - val_accuracy: 0.6516\n",
      "Epoch 47/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6007 - accuracy: 0.6748 - val_loss: 0.6078 - val_accuracy: 0.6683\n",
      "Epoch 48/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.6050 - accuracy: 0.6712 - val_loss: 0.6019 - val_accuracy: 0.6756\n",
      "Epoch 49/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5946 - accuracy: 0.6825 - val_loss: 0.5946 - val_accuracy: 0.6838\n",
      "Epoch 50/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5916 - accuracy: 0.6857 - val_loss: 0.6378 - val_accuracy: 0.6557\n",
      "Epoch 51/100\n",
      "44800/44800 [==============================] - 0s 4us/sample - loss: 0.5952 - accuracy: 0.6832 - val_loss: 0.5978 - val_accuracy: 0.6813\n",
      "Epoch 52/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5917 - accuracy: 0.6864 - val_loss: 0.6145 - val_accuracy: 0.6713\n",
      "Epoch 53/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5873 - accuracy: 0.6903 - val_loss: 0.5945 - val_accuracy: 0.6846\n",
      "Epoch 54/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5828 - accuracy: 0.6988 - val_loss: 0.5899 - val_accuracy: 0.6910\n",
      "Epoch 55/100\n",
      "44800/44800 [==============================] - 0s 4us/sample - loss: 0.5903 - accuracy: 0.6880 - val_loss: 0.5875 - val_accuracy: 0.6946\n",
      "Epoch 56/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5805 - accuracy: 0.6992 - val_loss: 0.5822 - val_accuracy: 0.7040\n",
      "Epoch 57/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5933 - accuracy: 0.6854 - val_loss: 0.6054 - val_accuracy: 0.6741\n",
      "Epoch 58/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5836 - accuracy: 0.6976 - val_loss: 0.6018 - val_accuracy: 0.6837\n",
      "Epoch 59/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5807 - accuracy: 0.6996 - val_loss: 0.5836 - val_accuracy: 0.6986\n",
      "Epoch 60/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5763 - accuracy: 0.7065 - val_loss: 0.5762 - val_accuracy: 0.7065\n",
      "Epoch 61/100\n",
      "44800/44800 [==============================] - 0s 4us/sample - loss: 0.5765 - accuracy: 0.7033 - val_loss: 0.5731 - val_accuracy: 0.7119\n",
      "Epoch 62/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5738 - accuracy: 0.7067 - val_loss: 0.5820 - val_accuracy: 0.6990\n",
      "Epoch 63/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5777 - accuracy: 0.7027 - val_loss: 0.5781 - val_accuracy: 0.7079\n",
      "Epoch 64/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5767 - accuracy: 0.7027 - val_loss: 0.5947 - val_accuracy: 0.6834\n",
      "Epoch 65/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5681 - accuracy: 0.7137 - val_loss: 0.5765 - val_accuracy: 0.7113\n",
      "Epoch 66/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5719 - accuracy: 0.7081 - val_loss: 0.5719 - val_accuracy: 0.7107\n",
      "Epoch 67/100\n",
      "44800/44800 [==============================] - 0s 4us/sample - loss: 0.5656 - accuracy: 0.7158 - val_loss: 0.5690 - val_accuracy: 0.7156\n",
      "Epoch 68/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5680 - accuracy: 0.7129 - val_loss: 0.5861 - val_accuracy: 0.6972\n",
      "Epoch 69/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5679 - accuracy: 0.7119 - val_loss: 0.5755 - val_accuracy: 0.7111\n",
      "Epoch 70/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5677 - accuracy: 0.7116 - val_loss: 0.5706 - val_accuracy: 0.7155\n",
      "Epoch 71/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5680 - accuracy: 0.7123 - val_loss: 0.5740 - val_accuracy: 0.7087\n",
      "Epoch 72/100\n",
      "44800/44800 [==============================] - 0s 4us/sample - loss: 0.5678 - accuracy: 0.7110 - val_loss: 0.5739 - val_accuracy: 0.7122\n",
      "Epoch 73/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5665 - accuracy: 0.7137 - val_loss: 0.5705 - val_accuracy: 0.7135\n",
      "Epoch 74/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5664 - accuracy: 0.7127 - val_loss: 0.5702 - val_accuracy: 0.7099\n",
      "Epoch 75/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5603 - accuracy: 0.7205 - val_loss: 0.5695 - val_accuracy: 0.7112\n",
      "Epoch 76/100\n",
      "44800/44800 [==============================] - 0s 6us/sample - loss: 0.5595 - accuracy: 0.7193 - val_loss: 0.5706 - val_accuracy: 0.7113\n",
      "Epoch 77/100\n",
      "44800/44800 [==============================] - 0s 6us/sample - loss: 0.5642 - accuracy: 0.7141 - val_loss: 0.5915 - val_accuracy: 0.6919\n",
      "Epoch 78/100\n",
      "44800/44800 [==============================] - 0s 6us/sample - loss: 0.5639 - accuracy: 0.7146 - val_loss: 0.5822 - val_accuracy: 0.6956\n",
      "Epoch 79/100\n",
      "44800/44800 [==============================] - 0s 7us/sample - loss: 0.5670 - accuracy: 0.7129 - val_loss: 0.5719 - val_accuracy: 0.7138\n",
      "Epoch 80/100\n",
      "44800/44800 [==============================] - 0s 6us/sample - loss: 0.5601 - accuracy: 0.7185 - val_loss: 0.6054 - val_accuracy: 0.6794\n",
      "Epoch 81/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5665 - accuracy: 0.7115 - val_loss: 0.5679 - val_accuracy: 0.7144\n",
      "Epoch 82/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5564 - accuracy: 0.7216 - val_loss: 0.5760 - val_accuracy: 0.7050\n",
      "Epoch 83/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5623 - accuracy: 0.7156 - val_loss: 0.5647 - val_accuracy: 0.7185\n",
      "Epoch 84/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5621 - accuracy: 0.7166 - val_loss: 0.5723 - val_accuracy: 0.7134\n",
      "Epoch 85/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5572 - accuracy: 0.7210 - val_loss: 0.5679 - val_accuracy: 0.7142\n",
      "Epoch 86/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5634 - accuracy: 0.7134 - val_loss: 0.5723 - val_accuracy: 0.7078\n",
      "Epoch 87/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5608 - accuracy: 0.7167 - val_loss: 0.5896 - val_accuracy: 0.6937\n",
      "Epoch 88/100\n",
      "44800/44800 [==============================] - 0s 6us/sample - loss: 0.5528 - accuracy: 0.7262 - val_loss: 0.5736 - val_accuracy: 0.7105\n",
      "Epoch 89/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5552 - accuracy: 0.7194 - val_loss: 0.5669 - val_accuracy: 0.7164\n",
      "Epoch 90/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5592 - accuracy: 0.7199 - val_loss: 0.5644 - val_accuracy: 0.7210\n",
      "Epoch 91/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5597 - accuracy: 0.7192 - val_loss: 0.5649 - val_accuracy: 0.7191\n",
      "Epoch 92/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5572 - accuracy: 0.7229 - val_loss: 0.5608 - val_accuracy: 0.7213\n",
      "Epoch 93/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5538 - accuracy: 0.7237 - val_loss: 0.5726 - val_accuracy: 0.7124\n",
      "Epoch 94/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5594 - accuracy: 0.7174 - val_loss: 0.5827 - val_accuracy: 0.7049\n",
      "Epoch 95/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5536 - accuracy: 0.7235 - val_loss: 0.5617 - val_accuracy: 0.7171\n",
      "Epoch 96/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5563 - accuracy: 0.7210 - val_loss: 0.5915 - val_accuracy: 0.6963\n",
      "Epoch 97/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5518 - accuracy: 0.7275 - val_loss: 0.5761 - val_accuracy: 0.7078\n",
      "Epoch 98/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5544 - accuracy: 0.7242 - val_loss: 0.5617 - val_accuracy: 0.7204\n",
      "Epoch 99/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5538 - accuracy: 0.7234 - val_loss: 0.5651 - val_accuracy: 0.7190\n",
      "Epoch 100/100\n",
      "44800/44800 [==============================] - 0s 5us/sample - loss: 0.5550 - accuracy: 0.7230 - val_loss: 0.5679 - val_accuracy: 0.7138\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "    positive       0.72      0.71      0.71      5621\n",
      "    negative       0.71      0.72      0.71      5579\n",
      "\n",
      "    accuracy                           0.71     11200\n",
      "   macro avg       0.71      0.71      0.71     11200\n",
      "weighted avg       0.71      0.71      0.71     11200\n",
      "\n",
      "Accuracy  : 0.7138392857142857\n",
      "Precision : 0.7139075785287651\n",
      "f1Score : 0.7138303544540637\n",
      "[[3975 1646]\n",
      " [1559 4020]]\n",
      "Results for fold 2\n",
      "Train on 44800 samples, validate on 11200 samples\n",
      "Epoch 1/100\n",
      "44800/44800 [==============================] - 0s 6us/sample - loss: 0.5617 - accuracy: 0.7160 - val_loss: 0.5507 - val_accuracy: 0.7252\n",
      "Epoch 2/100\n",
      "44800/44800 [==============================] - 0s 6us/sample - loss: 0.5561 - accuracy: 0.7233 - val_loss: 0.5549 - val_accuracy: 0.7237\n",
      "Epoch 3/100\n",
      "44800/44800 [==============================] - 0s 6us/sample - loss: 0.5540 - accuracy: 0.7235 - val_loss: 0.6127 - val_accuracy: 0.6676\n",
      "Epoch 4/100\n",
      "44800/44800 [==============================] - 0s 6us/sample - loss: 0.5610 - accuracy: 0.7163 - val_loss: 0.5511 - val_accuracy: 0.7262\n",
      "Epoch 5/100\n",
      "44800/44800 [==============================] - 0s 6us/sample - loss: 0.5525 - accuracy: 0.7257 - val_loss: 0.5618 - val_accuracy: 0.7194\n",
      "Epoch 6/100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "44800/44800 [==============================] - 1s 33us/sample - loss: 0.5576 - accuracy: 0.7209 - val_loss: 0.5727 - val_accuracy: 0.7057\n",
      "Epoch 7/100\n",
      "44800/44800 [==============================] - 1s 29us/sample - loss: 0.5603 - accuracy: 0.7174 - val_loss: 0.5508 - val_accuracy: 0.7287\n",
      "Epoch 8/100\n",
      "44800/44800 [==============================] - 1s 31us/sample - loss: 0.5536 - accuracy: 0.7233 - val_loss: 0.5540 - val_accuracy: 0.7244\n",
      "Epoch 9/100\n",
      "44800/44800 [==============================] - 1s 33us/sample - loss: 0.5506 - accuracy: 0.7276 - val_loss: 0.5705 - val_accuracy: 0.7080\n",
      "Epoch 10/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5578 - accuracy: 0.7200 - val_loss: 0.5628 - val_accuracy: 0.7176\n",
      "Epoch 11/100\n",
      "44800/44800 [==============================] - 1s 30us/sample - loss: 0.5540 - accuracy: 0.7236 - val_loss: 0.5733 - val_accuracy: 0.7018\n",
      "Epoch 12/100\n",
      "44800/44800 [==============================] - 1s 33us/sample - loss: 0.5490 - accuracy: 0.7287 - val_loss: 0.5586 - val_accuracy: 0.7204ss: 0.5494 - accuracy:  - ETA: 0s - loss: 0.5493 - ac\n",
      "Epoch 13/100\n",
      "44800/44800 [==============================] - 1s 33us/sample - loss: 0.5588 - accuracy: 0.7191 - val_loss: 0.5511 - val_accuracy: 0.7271\n",
      "Epoch 14/100\n",
      "44800/44800 [==============================] - 1s 31us/sample - loss: 0.5489 - accuracy: 0.7277 - val_loss: 0.5577 - val_accuracy: 0.7186\n",
      "Epoch 15/100\n",
      "44800/44800 [==============================] - 1s 30us/sample - loss: 0.5628 - accuracy: 0.7150 - val_loss: 0.6093 - val_accuracy: 0.6620\n",
      "Epoch 16/100\n",
      "44800/44800 [==============================] - 1s 29us/sample - loss: 0.5693 - accuracy: 0.7079 - val_loss: 0.5609 - val_accuracy: 0.7180\n",
      "Epoch 17/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5495 - accuracy: 0.7282 - val_loss: 0.5545 - val_accuracy: 0.7254\n",
      "Epoch 18/100\n",
      "44800/44800 [==============================] - 1s 27us/sample - loss: 0.5514 - accuracy: 0.7243 - val_loss: 0.5532 - val_accuracy: 0.7265\n",
      "Epoch 19/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5473 - accuracy: 0.7287 - val_loss: 0.6016 - val_accuracy: 0.6693\n",
      "Epoch 20/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5588 - accuracy: 0.7182 - val_loss: 0.6054 - val_accuracy: 0.6645\n",
      "Epoch 21/100\n",
      "44800/44800 [==============================] - 1s 22us/sample - loss: 0.5600 - accuracy: 0.7167 - val_loss: 0.5926 - val_accuracy: 0.6827\n",
      "Epoch 22/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5609 - accuracy: 0.7168 - val_loss: 0.5688 - val_accuracy: 0.7104\n",
      "Epoch 23/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5477 - accuracy: 0.7281 - val_loss: 0.5552 - val_accuracy: 0.7235\n",
      "Epoch 24/100\n",
      "44800/44800 [==============================] - ETA: 0s - loss: 0.5527 - accuracy: 0.72 - 1s 27us/sample - loss: 0.5534 - accuracy: 0.7231 - val_loss: 0.5829 - val_accuracy: 0.6946\n",
      "Epoch 25/100\n",
      "44800/44800 [==============================] - 1s 31us/sample - loss: 0.5576 - accuracy: 0.7199 - val_loss: 0.5511 - val_accuracy: 0.7257\n",
      "Epoch 26/100\n",
      "44800/44800 [==============================] - 2s 34us/sample - loss: 0.5538 - accuracy: 0.7232 - val_loss: 0.5547 - val_accuracy: 0.7251\n",
      "Epoch 27/100\n",
      "44800/44800 [==============================] - 2s 37us/sample - loss: 0.5494 - accuracy: 0.7292 - val_loss: 0.5496 - val_accuracy: 0.7271s - loss: 0.5511 - accuracy: 0.72 - ETA: 0s - loss: 0.5499 - accu - ETA: 0s - loss: 0.5495 - accuracy: 0.72\n",
      "Epoch 28/100\n",
      "44800/44800 [==============================] - 2s 34us/sample - loss: 0.5532 - accuracy: 0.7230 - val_loss: 0.5542 - val_accuracy: 0.7249\n",
      "Epoch 29/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5527 - accuracy: 0.7230 - val_loss: 0.5713 - val_accuracy: 0.7102\n",
      "Epoch 30/100\n",
      "44800/44800 [==============================] - 1s 29us/sample - loss: 0.5525 - accuracy: 0.7247 - val_loss: 0.5593 - val_accuracy: 0.7186\n",
      "Epoch 31/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5522 - accuracy: 0.7233 - val_loss: 0.5569 - val_accuracy: 0.7250\n",
      "Epoch 32/100\n",
      "44800/44800 [==============================] - ETA: 0s - loss: 0.5478 - accuracy: 0.72 - 1s 21us/sample - loss: 0.5487 - accuracy: 0.7271 - val_loss: 0.5765 - val_accuracy: 0.6985\n",
      "Epoch 33/100\n",
      "44800/44800 [==============================] - 1s 24us/sample - loss: 0.5475 - accuracy: 0.7283 - val_loss: 0.5539 - val_accuracy: 0.7238\n",
      "Epoch 34/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5488 - accuracy: 0.7280 - val_loss: 0.5611 - val_accuracy: 0.7202\n",
      "Epoch 35/100\n",
      "44800/44800 [==============================] - 1s 23us/sample - loss: 0.5527 - accuracy: 0.7234 - val_loss: 0.5819 - val_accuracy: 0.6965\n",
      "Epoch 36/100\n",
      "44800/44800 [==============================] - 1s 21us/sample - loss: 0.5581 - accuracy: 0.7171 - val_loss: 0.5514 - val_accuracy: 0.7284\n",
      "Epoch 37/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5506 - accuracy: 0.7250 - val_loss: 0.5690 - val_accuracy: 0.7120\n",
      "Epoch 38/100\n",
      "44800/44800 [==============================] - 1s 27us/sample - loss: 0.5557 - accuracy: 0.7211 - val_loss: 0.5525 - val_accuracy: 0.7249\n",
      "Epoch 39/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5569 - accuracy: 0.7194 - val_loss: 0.5555 - val_accuracy: 0.7229\n",
      "Epoch 40/100\n",
      "44800/44800 [==============================] - 1s 27us/sample - loss: 0.5552 - accuracy: 0.7210 - val_loss: 0.5527 - val_accuracy: 0.7268\n",
      "Epoch 41/100\n",
      "44800/44800 [==============================] - 1s 29us/sample - loss: 0.5507 - accuracy: 0.7262 - val_loss: 0.5553 - val_accuracy: 0.7255\n",
      "Epoch 42/100\n",
      "44800/44800 [==============================] - 1s 27us/sample - loss: 0.5524 - accuracy: 0.7243 - val_loss: 0.5523 - val_accuracy: 0.7283\n",
      "Epoch 43/100\n",
      "44800/44800 [==============================] - 1s 24us/sample - loss: 0.5468 - accuracy: 0.7297 - val_loss: 0.5712 - val_accuracy: 0.7132\n",
      "Epoch 44/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5507 - accuracy: 0.7261 - val_loss: 0.5515 - val_accuracy: 0.7286\n",
      "Epoch 45/100\n",
      "44800/44800 [==============================] - 1s 23us/sample - loss: 0.5465 - accuracy: 0.7305 - val_loss: 0.5567 - val_accuracy: 0.7224\n",
      "Epoch 46/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5498 - accuracy: 0.7265 - val_loss: 0.5656 - val_accuracy: 0.7079\n",
      "Epoch 47/100\n",
      "44800/44800 [==============================] - 1s 24us/sample - loss: 0.5494 - accuracy: 0.7260 - val_loss: 0.5560 - val_accuracy: 0.7258\n",
      "Epoch 48/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5509 - accuracy: 0.7249 - val_loss: 0.5600 - val_accuracy: 0.7186\n",
      "Epoch 49/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5518 - accuracy: 0.7226 - val_loss: 0.5524 - val_accuracy: 0.7254\n",
      "Epoch 50/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5483 - accuracy: 0.7270 - val_loss: 0.5720 - val_accuracy: 0.7070\n",
      "Epoch 51/100\n",
      "44800/44800 [==============================] - 1s 24us/sample - loss: 0.5495 - accuracy: 0.7276 - val_loss: 0.5572 - val_accuracy: 0.7204\n",
      "Epoch 52/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5471 - accuracy: 0.7288 - val_loss: 0.5673 - val_accuracy: 0.7122\n",
      "Epoch 53/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5513 - accuracy: 0.7233 - val_loss: 0.5577 - val_accuracy: 0.7231\n",
      "Epoch 54/100\n",
      "44800/44800 [==============================] - 1s 24us/sample - loss: 0.5484 - accuracy: 0.7263 - val_loss: 0.5510 - val_accuracy: 0.7286\n",
      "Epoch 55/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5490 - accuracy: 0.7269 - val_loss: 0.5605 - val_accuracy: 0.7195\n",
      "Epoch 56/100\n",
      "44800/44800 [==============================] - 1s 29us/sample - loss: 0.5450 - accuracy: 0.7318 - val_loss: 0.5540 - val_accuracy: 0.7272\n",
      "Epoch 57/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5477 - accuracy: 0.7262 - val_loss: 0.6017 - val_accuracy: 0.6695\n",
      "Epoch 58/100\n",
      "44800/44800 [==============================] - 1s 27us/sample - loss: 0.5629 - accuracy: 0.7129 - val_loss: 0.5534 - val_accuracy: 0.7248\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 59/100\n",
      "44800/44800 [==============================] - 1s 22us/sample - loss: 0.5466 - accuracy: 0.7298 - val_loss: 0.5553 - val_accuracy: 0.7238\n",
      "Epoch 60/100\n",
      "44800/44800 [==============================] - 1s 24us/sample - loss: 0.5499 - accuracy: 0.7271 - val_loss: 0.5530 - val_accuracy: 0.7269\n",
      "Epoch 61/100\n",
      "44800/44800 [==============================] - 1s 23us/sample - loss: 0.5466 - accuracy: 0.7302 - val_loss: 0.5548 - val_accuracy: 0.7239\n",
      "Epoch 62/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5450 - accuracy: 0.7295 - val_loss: 0.5598 - val_accuracy: 0.7197\n",
      "Epoch 63/100\n",
      "44800/44800 [==============================] - 1s 23us/sample - loss: 0.5482 - accuracy: 0.7277 - val_loss: 0.5515 - val_accuracy: 0.7274\n",
      "Epoch 64/100\n",
      "44800/44800 [==============================] - 1s 24us/sample - loss: 0.5571 - accuracy: 0.7171 - val_loss: 0.5803 - val_accuracy: 0.7004\n",
      "Epoch 65/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5475 - accuracy: 0.7283 - val_loss: 0.5528 - val_accuracy: 0.7263\n",
      "Epoch 66/100\n",
      "44800/44800 [==============================] - 1s 27us/sample - loss: 0.5462 - accuracy: 0.7300 - val_loss: 0.5590 - val_accuracy: 0.7196\n",
      "Epoch 67/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5496 - accuracy: 0.7273 - val_loss: 0.5514 - val_accuracy: 0.7269\n",
      "Epoch 68/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5492 - accuracy: 0.7259 - val_loss: 0.5516 - val_accuracy: 0.7304\n",
      "Epoch 69/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5463 - accuracy: 0.7281 - val_loss: 0.5498 - val_accuracy: 0.7297\n",
      "Epoch 70/100\n",
      "44800/44800 [==============================] - 1s 30us/sample - loss: 0.5440 - accuracy: 0.7294 - val_loss: 0.5566 - val_accuracy: 0.7245\n",
      "Epoch 71/100\n",
      "44800/44800 [==============================] - 1s 22us/sample - loss: 0.5555 - accuracy: 0.7194 - val_loss: 0.5519 - val_accuracy: 0.7259\n",
      "Epoch 72/100\n",
      "44800/44800 [==============================] - 1s 27us/sample - loss: 0.5478 - accuracy: 0.7278 - val_loss: 0.5533 - val_accuracy: 0.7252\n",
      "Epoch 73/100\n",
      "44800/44800 [==============================] - 1s 23us/sample - loss: 0.5425 - accuracy: 0.7321 - val_loss: 0.5567 - val_accuracy: 0.7227\n",
      "Epoch 74/100\n",
      "44800/44800 [==============================] - 1s 23us/sample - loss: 0.5489 - accuracy: 0.7271 - val_loss: 0.5567 - val_accuracy: 0.7190\n",
      "Epoch 75/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5436 - accuracy: 0.7321 - val_loss: 0.5664 - val_accuracy: 0.7149\n",
      "Epoch 76/100\n",
      "44800/44800 [==============================] - 1s 27us/sample - loss: 0.5418 - accuracy: 0.7322 - val_loss: 0.5599 - val_accuracy: 0.7197\n",
      "Epoch 77/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5544 - accuracy: 0.7201 - val_loss: 0.5532 - val_accuracy: 0.7256\n",
      "Epoch 78/100\n",
      "44800/44800 [==============================] - 1s 24us/sample - loss: 0.5514 - accuracy: 0.7225 - val_loss: 0.5565 - val_accuracy: 0.7251\n",
      "Epoch 79/100\n",
      "44800/44800 [==============================] - 1s 27us/sample - loss: 0.5433 - accuracy: 0.7320 - val_loss: 0.5502 - val_accuracy: 0.7296\n",
      "Epoch 80/100\n",
      "44800/44800 [==============================] - 1s 29us/sample - loss: 0.5494 - accuracy: 0.7250 - val_loss: 0.5806 - val_accuracy: 0.6989\n",
      "Epoch 81/100\n",
      "44800/44800 [==============================] - 2s 34us/sample - loss: 0.5509 - accuracy: 0.7240 - val_loss: 0.5525 - val_accuracy: 0.7262\n",
      "Epoch 82/100\n",
      "44800/44800 [==============================] - 1s 33us/sample - loss: 0.5458 - accuracy: 0.7300 - val_loss: 0.5509 - val_accuracy: 0.7267\n",
      "Epoch 83/100\n",
      "44800/44800 [==============================] - 1s 31us/sample - loss: 0.5412 - accuracy: 0.7339 - val_loss: 0.5765 - val_accuracy: 0.7086\n",
      "Epoch 84/100\n",
      "44800/44800 [==============================] - 1s 33us/sample - loss: 0.5553 - accuracy: 0.7194 - val_loss: 0.5649 - val_accuracy: 0.7184\n",
      "Epoch 85/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5440 - accuracy: 0.7314 - val_loss: 0.5737 - val_accuracy: 0.7046\n",
      "Epoch 86/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5522 - accuracy: 0.7224 - val_loss: 0.5557 - val_accuracy: 0.7244\n",
      "Epoch 87/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5452 - accuracy: 0.7304 - val_loss: 0.5510 - val_accuracy: 0.7276\n",
      "Epoch 88/100\n",
      "44800/44800 [==============================] - 1s 24us/sample - loss: 0.5424 - accuracy: 0.7327 - val_loss: 0.5583 - val_accuracy: 0.7192\n",
      "Epoch 89/100\n",
      "44800/44800 [==============================] - 1s 27us/sample - loss: 0.5482 - accuracy: 0.7273 - val_loss: 0.5857 - val_accuracy: 0.6962\n",
      "Epoch 90/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5450 - accuracy: 0.7300 - val_loss: 0.5521 - val_accuracy: 0.7281\n",
      "Epoch 91/100\n",
      "44800/44800 [==============================] - 1s 27us/sample - loss: 0.5442 - accuracy: 0.7286 - val_loss: 0.5523 - val_accuracy: 0.7294\n",
      "Epoch 92/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5440 - accuracy: 0.7313 - val_loss: 0.5666 - val_accuracy: 0.7142\n",
      "Epoch 93/100\n",
      "44800/44800 [==============================] - 1s 30us/sample - loss: 0.5418 - accuracy: 0.7326 - val_loss: 0.5776 - val_accuracy: 0.7042\n",
      "Epoch 94/100\n",
      "44800/44800 [==============================] - 1s 27us/sample - loss: 0.5462 - accuracy: 0.7283 - val_loss: 0.5505 - val_accuracy: 0.7294\n",
      "Epoch 95/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5474 - accuracy: 0.7302 - val_loss: 0.5727 - val_accuracy: 0.7071\n",
      "Epoch 96/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5535 - accuracy: 0.7202 - val_loss: 0.5621 - val_accuracy: 0.7209\n",
      "Epoch 97/100\n",
      "44800/44800 [==============================] - 1s 29us/sample - loss: 0.5406 - accuracy: 0.7329 - val_loss: 0.5623 - val_accuracy: 0.7212\n",
      "Epoch 98/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5488 - accuracy: 0.7269 - val_loss: 0.5568 - val_accuracy: 0.7220\n",
      "Epoch 99/100\n",
      "44800/44800 [==============================] - 1s 22us/sample - loss: 0.5466 - accuracy: 0.7292 - val_loss: 0.5748 - val_accuracy: 0.7092\n",
      "Epoch 100/100\n",
      "44800/44800 [==============================] - 1s 24us/sample - loss: 0.5451 - accuracy: 0.7306 - val_loss: 0.6035 - val_accuracy: 0.6741\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "    positive       0.62      0.89      0.73      5526\n",
      "    negative       0.81      0.47      0.59      5674\n",
      "\n",
      "    accuracy                           0.67     11200\n",
      "   macro avg       0.71      0.68      0.66     11200\n",
      "weighted avg       0.72      0.67      0.66     11200\n",
      "\n",
      "Accuracy  : 0.6741071428571429\n",
      "Precision : 0.7150647000308047\n",
      "f1Score : 0.6594475658436688\n",
      "[[4902  624]\n",
      " [3026 2648]]\n",
      "Results for fold 3\n",
      "Train on 44800 samples, validate on 11200 samples\n",
      "Epoch 1/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5702 - accuracy: 0.7057 - val_loss: 0.5508 - val_accuracy: 0.7202\n",
      "Epoch 2/100\n",
      "44800/44800 [==============================] - 1s 24us/sample - loss: 0.5588 - accuracy: 0.7176 - val_loss: 0.5286 - val_accuracy: 0.7439\n",
      "Epoch 3/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5546 - accuracy: 0.7221 - val_loss: 0.5320 - val_accuracy: 0.7456\n",
      "Epoch 4/100\n",
      "44800/44800 [==============================] - 1s 27us/sample - loss: 0.5501 - accuracy: 0.7252 - val_loss: 0.5350 - val_accuracy: 0.7378\n",
      "Epoch 5/100\n",
      "44800/44800 [==============================] - 1s 24us/sample - loss: 0.5577 - accuracy: 0.7186 - val_loss: 0.5414 - val_accuracy: 0.7388\n",
      "Epoch 6/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5455 - accuracy: 0.7301 - val_loss: 0.5321 - val_accuracy: 0.7447\n",
      "Epoch 7/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5471 - accuracy: 0.7276 - val_loss: 0.5465 - val_accuracy: 0.7274\n",
      "Epoch 8/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5537 - accuracy: 0.7211 - val_loss: 0.5396 - val_accuracy: 0.7322\n",
      "Epoch 9/100\n",
      "44800/44800 [==============================] - 1s 31us/sample - loss: 0.5472 - accuracy: 0.7281 - val_loss: 0.5344 - val_accuracy: 0.7412\n",
      "Epoch 10/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5520 - accuracy: 0.7240 - val_loss: 0.5391 - val_accuracy: 0.7363\n",
      "Epoch 11/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5451 - accuracy: 0.7298 - val_loss: 0.5347 - val_accuracy: 0.7416\n",
      "Epoch 12/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5512 - accuracy: 0.7238 - val_loss: 0.5569 - val_accuracy: 0.7200\n",
      "Epoch 13/100\n",
      "44800/44800 [==============================] - 1s 27us/sample - loss: 0.5541 - accuracy: 0.7228 - val_loss: 0.5346 - val_accuracy: 0.7440\n",
      "Epoch 14/100\n",
      "44800/44800 [==============================] - 1s 23us/sample - loss: 0.5448 - accuracy: 0.7283 - val_loss: 0.5323 - val_accuracy: 0.7431\n",
      "Epoch 15/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5505 - accuracy: 0.7239 - val_loss: 0.5469 - val_accuracy: 0.7245\n",
      "Epoch 16/100\n",
      "44800/44800 [==============================] - 1s 24us/sample - loss: 0.5534 - accuracy: 0.7213 - val_loss: 0.5478 - val_accuracy: 0.7266\n",
      "Epoch 17/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5485 - accuracy: 0.7269 - val_loss: 0.5319 - val_accuracy: 0.7446\n",
      "Epoch 18/100\n",
      "44800/44800 [==============================] - 1s 23us/sample - loss: 0.5482 - accuracy: 0.7271 - val_loss: 0.5682 - val_accuracy: 0.7048\n",
      "Epoch 19/100\n",
      "44800/44800 [==============================] - 1s 24us/sample - loss: 0.5486 - accuracy: 0.7258 - val_loss: 0.5400 - val_accuracy: 0.7379\n",
      "Epoch 20/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5467 - accuracy: 0.7289 - val_loss: 0.6161 - val_accuracy: 0.6635\n",
      "Epoch 21/100\n",
      "44800/44800 [==============================] - 1s 29us/sample - loss: 0.5589 - accuracy: 0.7187 - val_loss: 0.5435 - val_accuracy: 0.7314\n",
      "Epoch 22/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5473 - accuracy: 0.7285 - val_loss: 0.5563 - val_accuracy: 0.7224\n",
      "Epoch 23/100\n",
      "44800/44800 [==============================] - 1s 29us/sample - loss: 0.5555 - accuracy: 0.7193 - val_loss: 0.5397 - val_accuracy: 0.7379\n",
      "Epoch 24/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5532 - accuracy: 0.7205 - val_loss: 0.5575 - val_accuracy: 0.7187\n",
      "Epoch 25/100\n",
      "44800/44800 [==============================] - 1s 24us/sample - loss: 0.5516 - accuracy: 0.7252 - val_loss: 0.5379 - val_accuracy: 0.7386\n",
      "Epoch 26/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5464 - accuracy: 0.7282 - val_loss: 0.5343 - val_accuracy: 0.7404\n",
      "Epoch 27/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5429 - accuracy: 0.7317 - val_loss: 0.5368 - val_accuracy: 0.7372\n",
      "Epoch 28/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5445 - accuracy: 0.7315 - val_loss: 0.5435 - val_accuracy: 0.7316\n",
      "Epoch 29/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5472 - accuracy: 0.7272 - val_loss: 0.5345 - val_accuracy: 0.7384\n",
      "Epoch 30/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5454 - accuracy: 0.7289 - val_loss: 0.5337 - val_accuracy: 0.7404\n",
      "Epoch 31/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5475 - accuracy: 0.7278 - val_loss: 0.5588 - val_accuracy: 0.7162\n",
      "Epoch 32/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5505 - accuracy: 0.7252 - val_loss: 0.5814 - val_accuracy: 0.6980\n",
      "Epoch 33/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5493 - accuracy: 0.7253 - val_loss: 0.5423 - val_accuracy: 0.7355\n",
      "Epoch 34/100\n",
      "44800/44800 [==============================] - 1s 29us/sample - loss: 0.5424 - accuracy: 0.7321 - val_loss: 0.5357 - val_accuracy: 0.7366\n",
      "Epoch 35/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5409 - accuracy: 0.7313 - val_loss: 0.5321 - val_accuracy: 0.7433\n",
      "Epoch 36/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5505 - accuracy: 0.7253 - val_loss: 0.5343 - val_accuracy: 0.7440\n",
      "Epoch 37/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5413 - accuracy: 0.7326 - val_loss: 0.5677 - val_accuracy: 0.6898\n",
      "Epoch 38/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5486 - accuracy: 0.7250 - val_loss: 0.5467 - val_accuracy: 0.7282\n",
      "Epoch 39/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5462 - accuracy: 0.7283 - val_loss: 0.5390 - val_accuracy: 0.7388\n",
      "Epoch 40/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5435 - accuracy: 0.7308 - val_loss: 0.5445 - val_accuracy: 0.7299\n",
      "Epoch 41/100\n",
      "44800/44800 [==============================] - 1s 24us/sample - loss: 0.5444 - accuracy: 0.7298 - val_loss: 0.5462 - val_accuracy: 0.7338\n",
      "Epoch 42/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5433 - accuracy: 0.7301 - val_loss: 0.5692 - val_accuracy: 0.7032\n",
      "Epoch 43/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5528 - accuracy: 0.7237 - val_loss: 0.5376 - val_accuracy: 0.7415\n",
      "Epoch 44/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5468 - accuracy: 0.7278 - val_loss: 0.5396 - val_accuracy: 0.7396\n",
      "Epoch 45/100\n",
      "44800/44800 [==============================] - 1s 23us/sample - loss: 0.5467 - accuracy: 0.7283 - val_loss: 0.5380 - val_accuracy: 0.7379- loss: 0.5472 - accuracy\n",
      "Epoch 46/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5497 - accuracy: 0.7254 - val_loss: 0.5402 - val_accuracy: 0.7339\n",
      "Epoch 47/100\n",
      "44800/44800 [==============================] - 1s 29us/sample - loss: 0.5508 - accuracy: 0.7234 - val_loss: 0.5379 - val_accuracy: 0.7333\n",
      "Epoch 48/100\n",
      "44800/44800 [==============================] - 1s 29us/sample - loss: 0.5482 - accuracy: 0.7251 - val_loss: 0.5592 - val_accuracy: 0.7144\n",
      "Epoch 49/100\n",
      "44800/44800 [==============================] - 1s 29us/sample - loss: 0.5552 - accuracy: 0.7200 - val_loss: 0.5365 - val_accuracy: 0.7421\n",
      "Epoch 50/100\n",
      "44800/44800 [==============================] - 1s 27us/sample - loss: 0.5444 - accuracy: 0.7296 - val_loss: 0.5486 - val_accuracy: 0.7300\n",
      "Epoch 51/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5531 - accuracy: 0.7217 - val_loss: 0.5391 - val_accuracy: 0.7374\n",
      "Epoch 52/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5462 - accuracy: 0.7276 - val_loss: 0.5357 - val_accuracy: 0.7448\n",
      "Epoch 53/100\n",
      "44800/44800 [==============================] - ETA: 0s - loss: 0.5510 - accuracy: 0.72 - 1s 24us/sample - loss: 0.5509 - accuracy: 0.7234 - val_loss: 0.5346 - val_accuracy: 0.7421\n",
      "Epoch 54/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5419 - accuracy: 0.7322 - val_loss: 0.5393 - val_accuracy: 0.7393\n",
      "Epoch 55/100\n",
      "44800/44800 [==============================] - 1s 24us/sample - loss: 0.5576 - accuracy: 0.7188 - val_loss: 0.5822 - val_accuracy: 0.6927\n",
      "Epoch 56/100\n",
      "44800/44800 [==============================] - 1s 27us/sample - loss: 0.5478 - accuracy: 0.7265 - val_loss: 0.5343 - val_accuracy: 0.7415\n",
      "Epoch 57/100\n",
      "44800/44800 [==============================] - 1s 24us/sample - loss: 0.5418 - accuracy: 0.7318 - val_loss: 0.5350 - val_accuracy: 0.7415\n",
      "Epoch 58/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5410 - accuracy: 0.7331 - val_loss: 0.5325 - val_accuracy: 0.7445\n",
      "Epoch 59/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5413 - accuracy: 0.7317 - val_loss: 0.5508 - val_accuracy: 0.7301\n",
      "Epoch 60/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5473 - accuracy: 0.7264 - val_loss: 0.5548 - val_accuracy: 0.7200\n",
      "Epoch 61/100\n",
      "44800/44800 [==============================] - 1s 27us/sample - loss: 0.5431 - accuracy: 0.7310 - val_loss: 0.5360 - val_accuracy: 0.7422\n",
      "Epoch 62/100\n",
      "44800/44800 [==============================] - 1s 29us/sample - loss: 0.5466 - accuracy: 0.7282 - val_loss: 0.5433 - val_accuracy: 0.7309\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 63/100\n",
      "44800/44800 [==============================] - 1s 29us/sample - loss: 0.5421 - accuracy: 0.7299 - val_loss: 0.5329 - val_accuracy: 0.7425\n",
      "Epoch 64/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5432 - accuracy: 0.7300 - val_loss: 0.5487 - val_accuracy: 0.7261\n",
      "Epoch 65/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5440 - accuracy: 0.7298 - val_loss: 0.5486 - val_accuracy: 0.7285\n",
      "Epoch 66/100\n",
      "44800/44800 [==============================] - 1s 24us/sample - loss: 0.5449 - accuracy: 0.7296 - val_loss: 0.5497 - val_accuracy: 0.7276\n",
      "Epoch 67/100\n",
      "44800/44800 [==============================] - 1s 24us/sample - loss: 0.5438 - accuracy: 0.7300 - val_loss: 0.5419 - val_accuracy: 0.7338\n",
      "Epoch 68/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5466 - accuracy: 0.7283 - val_loss: 0.5482 - val_accuracy: 0.7284\n",
      "Epoch 69/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5506 - accuracy: 0.7231 - val_loss: 0.5381 - val_accuracy: 0.7392\n",
      "Epoch 70/100\n",
      "44800/44800 [==============================] - 1s 27us/sample - loss: 0.5427 - accuracy: 0.7312 - val_loss: 0.5384 - val_accuracy: 0.7336\n",
      "Epoch 71/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5512 - accuracy: 0.7233 - val_loss: 0.5357 - val_accuracy: 0.7405\n",
      "Epoch 72/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5462 - accuracy: 0.7271 - val_loss: 0.5351 - val_accuracy: 0.7436\n",
      "Epoch 73/100\n",
      "44800/44800 [==============================] - 1s 29us/sample - loss: 0.5401 - accuracy: 0.7320 - val_loss: 0.5491 - val_accuracy: 0.7272\n",
      "Epoch 74/100\n",
      "44800/44800 [==============================] - 2s 37us/sample - loss: 0.5602 - accuracy: 0.7144 - val_loss: 0.5373 - val_accuracy: 0.7379\n",
      "Epoch 75/100\n",
      "44800/44800 [==============================] - 2s 34us/sample - loss: 0.5437 - accuracy: 0.7301 - val_loss: 0.5350 - val_accuracy: 0.7406\n",
      "Epoch 76/100\n",
      "44800/44800 [==============================] - 1s 29us/sample - loss: 0.5424 - accuracy: 0.7315 - val_loss: 0.5326 - val_accuracy: 0.7416\n",
      "Epoch 77/100\n",
      "44800/44800 [==============================] - 2s 39us/sample - loss: 0.5491 - accuracy: 0.7250 - val_loss: 0.5366 - val_accuracy: 0.7352\n",
      "Epoch 78/100\n",
      "44800/44800 [==============================] - 2s 40us/sample - loss: 0.5424 - accuracy: 0.7310 - val_loss: 0.5403 - val_accuracy: 0.7377\n",
      "Epoch 79/100\n",
      "44800/44800 [==============================] - 1s 32us/sample - loss: 0.5443 - accuracy: 0.7297 - val_loss: 0.5353 - val_accuracy: 0.7415\n",
      "Epoch 80/100\n",
      "44800/44800 [==============================] - 1s 31us/sample - loss: 0.5495 - accuracy: 0.7230 - val_loss: 0.5399 - val_accuracy: 0.7406\n",
      "Epoch 81/100\n",
      "44800/44800 [==============================] - 2s 37us/sample - loss: 0.5496 - accuracy: 0.7244 - val_loss: 0.5606 - val_accuracy: 0.7195\n",
      "Epoch 82/100\n",
      "44800/44800 [==============================] - 1s 32us/sample - loss: 0.5594 - accuracy: 0.7142 - val_loss: 0.5513 - val_accuracy: 0.7196\n",
      "Epoch 83/100\n",
      "44800/44800 [==============================] - 2s 41us/sample - loss: 0.5458 - accuracy: 0.7277 - val_loss: 0.5578 - val_accuracy: 0.7157\n",
      "Epoch 84/100\n",
      "44800/44800 [==============================] - 2s 42us/sample - loss: 0.5494 - accuracy: 0.7254 - val_loss: 0.5391 - val_accuracy: 0.7405\n",
      "Epoch 85/100\n",
      "44800/44800 [==============================] - 2s 49us/sample - loss: 0.5470 - accuracy: 0.7256 - val_loss: 0.5477 - val_accuracy: 0.7339\n",
      "Epoch 86/100\n",
      "44800/44800 [==============================] - 2s 48us/sample - loss: 0.5443 - accuracy: 0.7300 - val_loss: 0.5380 - val_accuracy: 0.7357ETA: 0s - loss: 0.5443 - accu\n",
      "Epoch 87/100\n",
      "44800/44800 [==============================] - 2s 44us/sample - loss: 0.5424 - accuracy: 0.7313 - val_loss: 0.5484 - val_accuracy: 0.7311\n",
      "Epoch 88/100\n",
      "44800/44800 [==============================] - 2s 48us/sample - loss: 0.5399 - accuracy: 0.7345 - val_loss: 0.5324 - val_accuracy: 0.7419\n",
      "Epoch 89/100\n",
      "44800/44800 [==============================] - 2s 41us/sample - loss: 0.5421 - accuracy: 0.7304 - val_loss: 0.5642 - val_accuracy: 0.7171\n",
      "Epoch 90/100\n",
      "44800/44800 [==============================] - 2s 38us/sample - loss: 0.5431 - accuracy: 0.7296 - val_loss: 0.5461 - val_accuracy: 0.7299\n",
      "Epoch 91/100\n",
      "44800/44800 [==============================] - 2s 39us/sample - loss: 0.5433 - accuracy: 0.7290 - val_loss: 0.5364 - val_accuracy: 0.7385\n",
      "Epoch 92/100\n",
      "44800/44800 [==============================] - 2s 47us/sample - loss: 0.5475 - accuracy: 0.7270 - val_loss: 0.5461 - val_accuracy: 0.7277\n",
      "Epoch 93/100\n",
      "44800/44800 [==============================] - 2s 51us/sample - loss: 0.5392 - accuracy: 0.7340 - val_loss: 0.5335 - val_accuracy: 0.7418\n",
      "Epoch 94/100\n",
      "44800/44800 [==============================] - 2s 40us/sample - loss: 0.5392 - accuracy: 0.7338 - val_loss: 0.5353 - val_accuracy: 0.7420\n",
      "Epoch 95/100\n",
      "44800/44800 [==============================] - 2s 41us/sample - loss: 0.5588 - accuracy: 0.7161 - val_loss: 0.5399 - val_accuracy: 0.7342\n",
      "Epoch 96/100\n",
      "44800/44800 [==============================] - 2s 45us/sample - loss: 0.5426 - accuracy: 0.7293 - val_loss: 0.5360 - val_accuracy: 0.7415\n",
      "Epoch 97/100\n",
      "44800/44800 [==============================] - 2s 50us/sample - loss: 0.5433 - accuracy: 0.7292 - val_loss: 0.5467 - val_accuracy: 0.7310\n",
      "Epoch 98/100\n",
      "44800/44800 [==============================] - 2s 48us/sample - loss: 0.5434 - accuracy: 0.7306 - val_loss: 0.5376 - val_accuracy: 0.7400\n",
      "Epoch 99/100\n",
      "44800/44800 [==============================] - 2s 45us/sample - loss: 0.5381 - accuracy: 0.7338 - val_loss: 0.5414 - val_accuracy: 0.7359\n",
      "Epoch 100/100\n",
      "44800/44800 [==============================] - 2s 46us/sample - loss: 0.5389 - accuracy: 0.7335 - val_loss: 0.5580 - val_accuracy: 0.7141\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "    positive       0.77      0.61      0.68      5560\n",
      "    negative       0.68      0.82      0.74      5640\n",
      "\n",
      "    accuracy                           0.71     11200\n",
      "   macro avg       0.72      0.71      0.71     11200\n",
      "weighted avg       0.72      0.71      0.71     11200\n",
      "\n",
      "Accuracy  : 0.7141071428571428\n",
      "Precision : 0.7236702034082737\n",
      "f1Score : 0.7106636959036666\n",
      "[[3368 2192]\n",
      " [1010 4630]]\n",
      "Results for fold 4\n",
      "Train on 44800 samples, validate on 11200 samples\n",
      "Epoch 1/100\n",
      "44800/44800 [==============================] - 2s 38us/sample - loss: 0.5394 - accuracy: 0.7353 - val_loss: 0.5333 - val_accuracy: 0.7381\n",
      "Epoch 2/100\n",
      "44800/44800 [==============================] - 1s 33us/sample - loss: 0.5452 - accuracy: 0.7294 - val_loss: 0.5447 - val_accuracy: 0.7262\n",
      "Epoch 3/100\n",
      "44800/44800 [==============================] - 1s 32us/sample - loss: 0.5465 - accuracy: 0.7284 - val_loss: 0.5405 - val_accuracy: 0.7297\n",
      "Epoch 4/100\n",
      "44800/44800 [==============================] - 2s 37us/sample - loss: 0.5463 - accuracy: 0.7287 - val_loss: 0.5437 - val_accuracy: 0.7281\n",
      "Epoch 5/100\n",
      "44800/44800 [==============================] - 1s 31us/sample - loss: 0.5471 - accuracy: 0.7256 - val_loss: 0.5520 - val_accuracy: 0.7186\n",
      "Epoch 6/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5408 - accuracy: 0.7319 - val_loss: 0.5389 - val_accuracy: 0.7321\n",
      "Epoch 7/100\n",
      "44800/44800 [==============================] - 1s 29us/sample - loss: 0.5397 - accuracy: 0.7336 - val_loss: 0.5986 - val_accuracy: 0.6727\n",
      "Epoch 8/100\n",
      "44800/44800 [==============================] - 1s 30us/sample - loss: 0.5457 - accuracy: 0.7285 - val_loss: 0.5480 - val_accuracy: 0.7236\n",
      "Epoch 9/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5431 - accuracy: 0.7298 - val_loss: 0.5445 - val_accuracy: 0.7290\n",
      "Epoch 10/100\n",
      "44800/44800 [==============================] - 2s 36us/sample - loss: 0.5415 - accuracy: 0.7315 - val_loss: 0.5458 - val_accuracy: 0.7254\n",
      "Epoch 11/100\n",
      "44800/44800 [==============================] - 2s 41us/sample - loss: 0.5398 - accuracy: 0.7332 - val_loss: 0.5565 - val_accuracy: 0.7121\n",
      "Epoch 12/100\n",
      "44800/44800 [==============================] - 2s 46us/sample - loss: 0.5412 - accuracy: 0.7326 - val_loss: 0.5565 - val_accuracy: 0.7155\n",
      "Epoch 13/100\n",
      "44800/44800 [==============================] - 2s 38us/sample - loss: 0.5444 - accuracy: 0.7290 - val_loss: 0.5367 - val_accuracy: 0.7357\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 14/100\n",
      "44800/44800 [==============================] - 2s 38us/sample - loss: 0.5410 - accuracy: 0.7324 - val_loss: 0.5434 - val_accuracy: 0.7262\n",
      "Epoch 15/100\n",
      "44800/44800 [==============================] - 2s 42us/sample - loss: 0.5366 - accuracy: 0.7360 - val_loss: 0.5403 - val_accuracy: 0.7329\n",
      "Epoch 16/100\n",
      "44800/44800 [==============================] - 2s 43us/sample - loss: 0.5402 - accuracy: 0.7334 - val_loss: 0.5388 - val_accuracy: 0.7345\n",
      "Epoch 17/100\n",
      "44800/44800 [==============================] - 2s 40us/sample - loss: 0.5430 - accuracy: 0.7312 - val_loss: 0.5522 - val_accuracy: 0.7188\n",
      "Epoch 18/100\n",
      "44800/44800 [==============================] - 1s 31us/sample - loss: 0.5372 - accuracy: 0.7356 - val_loss: 0.5794 - val_accuracy: 0.7000\n",
      "Epoch 19/100\n",
      "44800/44800 [==============================] - 1s 31us/sample - loss: 0.5366 - accuracy: 0.7356 - val_loss: 0.5398 - val_accuracy: 0.7310\n",
      "Epoch 20/100\n",
      "44800/44800 [==============================] - 1s 31us/sample - loss: 0.5419 - accuracy: 0.7329 - val_loss: 0.5498 - val_accuracy: 0.7255\n",
      "Epoch 21/100\n",
      "44800/44800 [==============================] - 2s 36us/sample - loss: 0.5418 - accuracy: 0.7307 - val_loss: 0.5440 - val_accuracy: 0.7275\n",
      "Epoch 22/100\n",
      "44800/44800 [==============================] - 2s 45us/sample - loss: 0.5375 - accuracy: 0.7346 - val_loss: 0.5455 - val_accuracy: 0.7277\n",
      "Epoch 23/100\n",
      "44800/44800 [==============================] - 2s 48us/sample - loss: 0.5425 - accuracy: 0.7305 - val_loss: 0.5512 - val_accuracy: 0.7242\n",
      "Epoch 24/100\n",
      "44800/44800 [==============================] - 2s 47us/sample - loss: 0.5376 - accuracy: 0.7336 - val_loss: 0.5419 - val_accuracy: 0.7328\n",
      "Epoch 25/100\n",
      "44800/44800 [==============================] - 1s 32us/sample - loss: 0.5403 - accuracy: 0.7343 - val_loss: 0.5402 - val_accuracy: 0.7367\n",
      "Epoch 26/100\n",
      "44800/44800 [==============================] - 1s 27us/sample - loss: 0.5394 - accuracy: 0.7332 - val_loss: 0.5382 - val_accuracy: 0.7365\n",
      "Epoch 27/100\n",
      "44800/44800 [==============================] - 1s 27us/sample - loss: 0.5375 - accuracy: 0.7350 - val_loss: 0.5393 - val_accuracy: 0.7337\n",
      "Epoch 28/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5358 - accuracy: 0.7372 - val_loss: 0.5397 - val_accuracy: 0.7330\n",
      "Epoch 29/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5358 - accuracy: 0.7356 - val_loss: 0.5480 - val_accuracy: 0.7262\n",
      "Epoch 30/100\n",
      "44800/44800 [==============================] - 1s 30us/sample - loss: 0.5432 - accuracy: 0.7301 - val_loss: 0.5415 - val_accuracy: 0.7319\n",
      "Epoch 31/100\n",
      "44800/44800 [==============================] - 1s 33us/sample - loss: 0.5405 - accuracy: 0.7318 - val_loss: 0.5419 - val_accuracy: 0.7339\n",
      "Epoch 32/100\n",
      "44800/44800 [==============================] - 2s 35us/sample - loss: 0.5370 - accuracy: 0.7341 - val_loss: 0.5405 - val_accuracy: 0.7324\n",
      "Epoch 33/100\n",
      "44800/44800 [==============================] - 2s 35us/sample - loss: 0.5369 - accuracy: 0.7362 - val_loss: 0.5420 - val_accuracy: 0.7335\n",
      "Epoch 34/100\n",
      "44800/44800 [==============================] - 1s 30us/sample - loss: 0.5396 - accuracy: 0.7336 - val_loss: 0.5740 - val_accuracy: 0.7106\n",
      "Epoch 35/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5496 - accuracy: 0.7246 - val_loss: 0.5504 - val_accuracy: 0.7224\n",
      "Epoch 36/100\n",
      "44800/44800 [==============================] - 1s 32us/sample - loss: 0.5345 - accuracy: 0.7370 - val_loss: 0.5409 - val_accuracy: 0.7342\n",
      "Epoch 37/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5410 - accuracy: 0.7310 - val_loss: 0.5481 - val_accuracy: 0.7252\n",
      "Epoch 38/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5366 - accuracy: 0.7355 - val_loss: 0.5569 - val_accuracy: 0.7178\n",
      "Epoch 39/100\n",
      "44800/44800 [==============================] - 1s 30us/sample - loss: 0.5409 - accuracy: 0.7330 - val_loss: 0.5464 - val_accuracy: 0.7277\n",
      "Epoch 40/100\n",
      "44800/44800 [==============================] - 1s 30us/sample - loss: 0.5443 - accuracy: 0.7276 - val_loss: 0.5418 - val_accuracy: 0.7331\n",
      "Epoch 41/100\n",
      "44800/44800 [==============================] - 2s 35us/sample - loss: 0.5336 - accuracy: 0.7384 - val_loss: 0.5414 - val_accuracy: 0.7361\n",
      "Epoch 42/100\n",
      "44800/44800 [==============================] - 2s 38us/sample - loss: 0.5381 - accuracy: 0.7343 - val_loss: 0.5512 - val_accuracy: 0.7214A: 0s - loss: 0.5346 - accura\n",
      "Epoch 43/100\n",
      "44800/44800 [==============================] - 2s 36us/sample - loss: 0.5407 - accuracy: 0.7331 - val_loss: 0.5435 - val_accuracy: 0.7329\n",
      "Epoch 44/100\n",
      "44800/44800 [==============================] - 2s 39us/sample - loss: 0.5340 - accuracy: 0.7372 - val_loss: 0.5439 - val_accuracy: 0.7285\n",
      "Epoch 45/100\n",
      "44800/44800 [==============================] - 2s 34us/sample - loss: 0.5399 - accuracy: 0.7331 - val_loss: 0.5412 - val_accuracy: 0.7322\n",
      "Epoch 46/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5360 - accuracy: 0.7358 - val_loss: 0.5435 - val_accuracy: 0.7329\n",
      "Epoch 47/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5376 - accuracy: 0.7354 - val_loss: 0.5439 - val_accuracy: 0.7326\n",
      "Epoch 48/100\n",
      "44800/44800 [==============================] - 1s 25us/sample - loss: 0.5355 - accuracy: 0.7358 - val_loss: 0.5486 - val_accuracy: 0.7290\n",
      "Epoch 49/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5341 - accuracy: 0.7376 - val_loss: 0.5586 - val_accuracy: 0.7224\n",
      "Epoch 50/100\n",
      "44800/44800 [==============================] - 2s 37us/sample - loss: 0.5532 - accuracy: 0.7207 - val_loss: 0.5679 - val_accuracy: 0.7077\n",
      "Epoch 51/100\n",
      "44800/44800 [==============================] - 1s 33us/sample - loss: 0.5424 - accuracy: 0.7304 - val_loss: 0.5491 - val_accuracy: 0.7262\n",
      "Epoch 52/100\n",
      "44800/44800 [==============================] - 2s 34us/sample - loss: 0.5372 - accuracy: 0.7331 - val_loss: 0.5468 - val_accuracy: 0.7315\n",
      "Epoch 53/100\n",
      "44800/44800 [==============================] - 2s 35us/sample - loss: 0.5328 - accuracy: 0.7393 - val_loss: 0.5435 - val_accuracy: 0.7353\n",
      "Epoch 54/100\n",
      "44800/44800 [==============================] - 2s 40us/sample - loss: 0.5395 - accuracy: 0.7333 - val_loss: 0.5702 - val_accuracy: 0.7086\n",
      "Epoch 55/100\n",
      "44800/44800 [==============================] - 2s 39us/sample - loss: 0.5467 - accuracy: 0.7275 - val_loss: 0.5654 - val_accuracy: 0.7129\n",
      "Epoch 56/100\n",
      "44800/44800 [==============================] - 2s 38us/sample - loss: 0.5365 - accuracy: 0.7363 - val_loss: 0.5438 - val_accuracy: 0.7336\n",
      "Epoch 57/100\n",
      "44800/44800 [==============================] - 1s 32us/sample - loss: 0.5338 - accuracy: 0.7370 - val_loss: 0.5421 - val_accuracy: 0.7334\n",
      "Epoch 58/100\n",
      "44800/44800 [==============================] - 2s 34us/sample - loss: 0.5357 - accuracy: 0.7367 - val_loss: 0.5491 - val_accuracy: 0.7251\n",
      "Epoch 59/100\n",
      "44800/44800 [==============================] - 2s 35us/sample - loss: 0.5383 - accuracy: 0.7339 - val_loss: 0.5470 - val_accuracy: 0.7311\n",
      "Epoch 60/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5341 - accuracy: 0.7370 - val_loss: 0.5469 - val_accuracy: 0.7306\n",
      "Epoch 61/100\n",
      "44800/44800 [==============================] - 1s 29us/sample - loss: 0.5327 - accuracy: 0.7393 - val_loss: 0.5429 - val_accuracy: 0.7334\n",
      "Epoch 62/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5405 - accuracy: 0.7309 - val_loss: 0.5445 - val_accuracy: 0.7325\n",
      "Epoch 63/100\n",
      "44800/44800 [==============================] - 1s 30us/sample - loss: 0.5409 - accuracy: 0.7313 - val_loss: 0.5465 - val_accuracy: 0.7282\n",
      "Epoch 64/100\n",
      "44800/44800 [==============================] - 2s 36us/sample - loss: 0.5314 - accuracy: 0.7393 - val_loss: 0.5418 - val_accuracy: 0.7337\n",
      "Epoch 65/100\n",
      "44800/44800 [==============================] - 2s 42us/sample - loss: 0.5332 - accuracy: 0.7367 - val_loss: 0.5424 - val_accuracy: 0.7337\n",
      "Epoch 66/100\n",
      "44800/44800 [==============================] - 2s 36us/sample - loss: 0.5369 - accuracy: 0.7351 - val_loss: 0.5492 - val_accuracy: 0.7266\n",
      "Epoch 67/100\n",
      "44800/44800 [==============================] - 2s 41us/sample - loss: 0.5318 - accuracy: 0.7385 - val_loss: 0.5875 - val_accuracy: 0.6890\n",
      "Epoch 68/100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "44800/44800 [==============================] - 2s 35us/sample - loss: 0.5349 - accuracy: 0.7362 - val_loss: 0.5477 - val_accuracy: 0.7295\n",
      "Epoch 69/100\n",
      "44800/44800 [==============================] - 1s 32us/sample - loss: 0.5392 - accuracy: 0.7334 - val_loss: 0.5753 - val_accuracy: 0.7021\n",
      "Epoch 70/100\n",
      "44800/44800 [==============================] - 2s 39us/sample - loss: 0.5355 - accuracy: 0.7357 - val_loss: 0.5613 - val_accuracy: 0.7153\n",
      "Epoch 71/100\n",
      "44800/44800 [==============================] - 1s 29us/sample - loss: 0.5464 - accuracy: 0.7265 - val_loss: 0.5548 - val_accuracy: 0.7233\n",
      "Epoch 72/100\n",
      "44800/44800 [==============================] - 1s 30us/sample - loss: 0.5402 - accuracy: 0.7329 - val_loss: 0.5566 - val_accuracy: 0.7251ss: 0.5395 - accuracy: 0.\n",
      "Epoch 73/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5357 - accuracy: 0.7358 - val_loss: 0.5717 - val_accuracy: 0.7063\n",
      "Epoch 74/100\n",
      "44800/44800 [==============================] - 1s 32us/sample - loss: 0.5366 - accuracy: 0.7356 - val_loss: 0.5502 - val_accuracy: 0.7281\n",
      "Epoch 75/100\n",
      "44800/44800 [==============================] - 2s 37us/sample - loss: 0.5340 - accuracy: 0.7376 - val_loss: 0.5482 - val_accuracy: 0.7323\n",
      "Epoch 76/100\n",
      "44800/44800 [==============================] - 2s 39us/sample - loss: 0.5421 - accuracy: 0.7299 - val_loss: 0.5579 - val_accuracy: 0.7178\n",
      "Epoch 77/100\n",
      "44800/44800 [==============================] - 2s 39us/sample - loss: 0.5341 - accuracy: 0.7373 - val_loss: 0.5425 - val_accuracy: 0.7340\n",
      "Epoch 78/100\n",
      "44800/44800 [==============================] - 1s 32us/sample - loss: 0.5316 - accuracy: 0.7383 - val_loss: 0.5433 - val_accuracy: 0.7329\n",
      "Epoch 79/100\n",
      "44800/44800 [==============================] - 2s 35us/sample - loss: 0.5336 - accuracy: 0.7375 - val_loss: 0.5513 - val_accuracy: 0.7240\n",
      "Epoch 80/100\n",
      "44800/44800 [==============================] - 2s 38us/sample - loss: 0.5305 - accuracy: 0.7391 - val_loss: 0.5424 - val_accuracy: 0.7354\n",
      "Epoch 81/100\n",
      "44800/44800 [==============================] - 2s 34us/sample - loss: 0.5313 - accuracy: 0.7394 - val_loss: 0.5434 - val_accuracy: 0.7338\n",
      "Epoch 82/100\n",
      "44800/44800 [==============================] - 1s 32us/sample - loss: 0.5338 - accuracy: 0.7369 - val_loss: 0.5466 - val_accuracy: 0.7317\n",
      "Epoch 83/100\n",
      "44800/44800 [==============================] - 1s 33us/sample - loss: 0.5365 - accuracy: 0.7349 - val_loss: 0.5448 - val_accuracy: 0.7346\n",
      "Epoch 84/100\n",
      "44800/44800 [==============================] - 2s 37us/sample - loss: 0.5354 - accuracy: 0.7364 - val_loss: 0.5479 - val_accuracy: 0.7287\n",
      "Epoch 85/100\n",
      "44800/44800 [==============================] - 1s 33us/sample - loss: 0.5323 - accuracy: 0.7369 - val_loss: 0.5605 - val_accuracy: 0.7206\n",
      "Epoch 86/100\n",
      "44800/44800 [==============================] - 1s 27us/sample - loss: 0.5319 - accuracy: 0.7391 - val_loss: 0.5462 - val_accuracy: 0.7288\n",
      "Epoch 87/100\n",
      "44800/44800 [==============================] - 1s 27us/sample - loss: 0.5341 - accuracy: 0.7382 - val_loss: 0.6006 - val_accuracy: 0.6756\n",
      "Epoch 88/100\n",
      "44800/44800 [==============================] - 1s 30us/sample - loss: 0.5382 - accuracy: 0.7334 - val_loss: 0.5466 - val_accuracy: 0.7321\n",
      "Epoch 89/100\n",
      "44800/44800 [==============================] - 1s 32us/sample - loss: 0.5328 - accuracy: 0.7383 - val_loss: 0.5520 - val_accuracy: 0.7256\n",
      "Epoch 90/100\n",
      "44800/44800 [==============================] - 2s 35us/sample - loss: 0.5430 - accuracy: 0.7298 - val_loss: 0.5686 - val_accuracy: 0.7110\n",
      "Epoch 91/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5333 - accuracy: 0.7387 - val_loss: 0.5563 - val_accuracy: 0.7209\n",
      "Epoch 92/100\n",
      "44800/44800 [==============================] - 1s 26us/sample - loss: 0.5352 - accuracy: 0.7360 - val_loss: 0.5660 - val_accuracy: 0.7144\n",
      "Epoch 93/100\n",
      "44800/44800 [==============================] - 2s 44us/sample - loss: 0.5365 - accuracy: 0.7345 - val_loss: 0.5522 - val_accuracy: 0.7271\n",
      "Epoch 94/100\n",
      "44800/44800 [==============================] - 2s 36us/sample - loss: 0.5353 - accuracy: 0.7371 - val_loss: 0.5566 - val_accuracy: 0.7209\n",
      "Epoch 95/100\n",
      "44800/44800 [==============================] - 1s 30us/sample - loss: 0.5297 - accuracy: 0.7408 - val_loss: 0.5550 - val_accuracy: 0.7228\n",
      "Epoch 96/100\n",
      "44800/44800 [==============================] - 1s 33us/sample - loss: 0.5423 - accuracy: 0.7296 - val_loss: 0.5666 - val_accuracy: 0.7132\n",
      "Epoch 97/100\n",
      "44800/44800 [==============================] - 2s 39us/sample - loss: 0.5441 - accuracy: 0.7280 - val_loss: 0.5501 - val_accuracy: 0.7268\n",
      "Epoch 98/100\n",
      "44800/44800 [==============================] - 2s 40us/sample - loss: 0.5342 - accuracy: 0.7359 - val_loss: 0.5483 - val_accuracy: 0.7304\n",
      "Epoch 99/100\n",
      "44800/44800 [==============================] - 1s 31us/sample - loss: 0.5338 - accuracy: 0.7371 - val_loss: 0.5445 - val_accuracy: 0.7330\n",
      "Epoch 100/100\n",
      "44800/44800 [==============================] - 2s 34us/sample - loss: 0.5327 - accuracy: 0.7379 - val_loss: 0.5624 - val_accuracy: 0.7181\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "    positive       0.75      0.67      0.71      5683\n",
      "    negative       0.69      0.76      0.73      5517\n",
      "\n",
      "    accuracy                           0.72     11200\n",
      "   macro avg       0.72      0.72      0.72     11200\n",
      "weighted avg       0.72      0.72      0.72     11200\n",
      "\n",
      "Accuracy  : 0.718125\n",
      "Precision : 0.7206144949460865\n",
      "f1Score : 0.7176603201713946\n",
      "[[3832 1851]\n",
      " [1306 4211]]\n",
      "Results for fold 5\n",
      "Train on 44800 samples, validate on 11200 samples\n",
      "Epoch 1/100\n",
      "44800/44800 [==============================] - 1s 31us/sample - loss: 0.5362 - accuracy: 0.7365 - val_loss: 0.5369 - val_accuracy: 0.7349\n",
      "Epoch 2/100\n",
      "44800/44800 [==============================] - 1s 32us/sample - loss: 0.5416 - accuracy: 0.7324 - val_loss: 0.5299 - val_accuracy: 0.7419\n",
      "Epoch 3/100\n",
      "44800/44800 [==============================] - 2s 36us/sample - loss: 0.5416 - accuracy: 0.7308 - val_loss: 0.5329 - val_accuracy: 0.7408\n",
      "Epoch 4/100\n",
      "44800/44800 [==============================] - 1s 33us/sample - loss: 0.5354 - accuracy: 0.7365 - val_loss: 0.5444 - val_accuracy: 0.7226\n",
      "Epoch 5/100\n",
      "44800/44800 [==============================] - 1s 32us/sample - loss: 0.5440 - accuracy: 0.7283 - val_loss: 0.5285 - val_accuracy: 0.7395\n",
      "Epoch 6/100\n",
      "44800/44800 [==============================] - 2s 43us/sample - loss: 0.5358 - accuracy: 0.7363 - val_loss: 0.5305 - val_accuracy: 0.7442\n",
      "Epoch 7/100\n",
      "44800/44800 [==============================] - 2s 36us/sample - loss: 0.5400 - accuracy: 0.7334 - val_loss: 0.5381 - val_accuracy: 0.7339\n",
      "Epoch 8/100\n",
      "44800/44800 [==============================] - 2s 36us/sample - loss: 0.5411 - accuracy: 0.7319 - val_loss: 0.5428 - val_accuracy: 0.7277\n",
      "Epoch 9/100\n",
      "44800/44800 [==============================] - 1s 32us/sample - loss: 0.5333 - accuracy: 0.7368 - val_loss: 0.5371 - val_accuracy: 0.7346\n",
      "Epoch 10/100\n",
      "44800/44800 [==============================] - 1s 28us/sample - loss: 0.5378 - accuracy: 0.7338 - val_loss: 0.5784 - val_accuracy: 0.6935\n",
      "Epoch 11/100\n",
      "44800/44800 [==============================] - 1s 30us/sample - loss: 0.5384 - accuracy: 0.7347 - val_loss: 0.5471 - val_accuracy: 0.7215\n",
      "Epoch 12/100\n",
      "44800/44800 [==============================] - 2s 38us/sample - loss: 0.5347 - accuracy: 0.7387 - val_loss: 0.5478 - val_accuracy: 0.719738 - accuracy: \n",
      "Epoch 13/100\n",
      "44800/44800 [==============================] - 2s 38us/sample - loss: 0.5357 - accuracy: 0.7356 - val_loss: 0.5566 - val_accuracy: 0.7198\n",
      "Epoch 14/100\n",
      "44800/44800 [==============================] - 2s 45us/sample - loss: 0.5395 - accuracy: 0.7338 - val_loss: 0.5633 - val_accuracy: 0.7121\n",
      "Epoch 15/100\n",
      "44800/44800 [==============================] - 2s 47us/sample - loss: 0.5404 - accuracy: 0.7309 - val_loss: 0.5539 - val_accuracy: 0.7168\n",
      "Epoch 16/100\n",
      "44800/44800 [==============================] - 2s 44us/sample - loss: 0.5331 - accuracy: 0.7367 - val_loss: 0.5365 - val_accuracy: 0.7376\n",
      "Epoch 17/100\n",
      "44800/44800 [==============================] - 2s 43us/sample - loss: 0.5351 - accuracy: 0.7368 - val_loss: 0.5326 - val_accuracy: 0.7399\n",
      "Epoch 18/100\n",
      "44800/44800 [==============================] - 2s 44us/sample - loss: 0.5326 - accuracy: 0.7391 - val_loss: 0.5479 - val_accuracy: 0.725915 - accuracy - ETA: 0s - loss: 0.5313 - accuracy: \n",
      "Epoch 19/100\n",
      "44800/44800 [==============================] - 2s 34us/sample - loss: 0.5339 - accuracy: 0.7375 - val_loss: 0.5325 - val_accuracy: 0.7399\n",
      "Epoch 20/100\n",
      "44800/44800 [==============================] - 2s 48us/sample - loss: 0.5346 - accuracy: 0.7363 - val_loss: 0.5336 - val_accuracy: 0.7387\n",
      "Epoch 21/100\n",
      "44800/44800 [==============================] - 2s 35us/sample - loss: 0.5332 - accuracy: 0.7381 - val_loss: 0.5340 - val_accuracy: 0.7369\n",
      "Epoch 22/100\n",
      "44800/44800 [==============================] - 1s 32us/sample - loss: 0.5426 - accuracy: 0.7299 - val_loss: 0.5399 - val_accuracy: 0.7348\n",
      "Epoch 23/100\n",
      "44800/44800 [==============================] - 2s 43us/sample - loss: 0.5333 - accuracy: 0.7390 - val_loss: 0.5386 - val_accuracy: 0.7376\n",
      "Epoch 24/100\n",
      "44800/44800 [==============================] - 2s 44us/sample - loss: 0.5338 - accuracy: 0.7387 - val_loss: 0.5455 - val_accuracy: 0.7248\n",
      "Epoch 25/100\n",
      "44800/44800 [==============================] - 3s 57us/sample - loss: 0.5381 - accuracy: 0.7326 - val_loss: 0.5900 - val_accuracy: 0.6854- loss: 0.5328 - accura\n",
      "Epoch 26/100\n",
      "44800/44800 [==============================] - 2s 55us/sample - loss: 0.5398 - accuracy: 0.7324 - val_loss: 0.5349 - val_accuracy: 0.7372\n",
      "Epoch 27/100\n",
      "44800/44800 [==============================] - 2s 42us/sample - loss: 0.5385 - accuracy: 0.7329 - val_loss: 0.5341 - val_accuracy: 0.7384\n",
      "Epoch 28/100\n",
      "44800/44800 [==============================] - 1s 32us/sample - loss: 0.5325 - accuracy: 0.7382 - val_loss: 0.5427 - val_accuracy: 0.7309\n",
      "Epoch 29/100\n",
      "44800/44800 [==============================] - 2s 40us/sample - loss: 0.5332 - accuracy: 0.7370 - val_loss: 0.5313 - val_accuracy: 0.7401\n",
      "Epoch 30/100\n",
      "44800/44800 [==============================] - 1s 31us/sample - loss: 0.5334 - accuracy: 0.7365 - val_loss: 0.5377 - val_accuracy: 0.7379\n",
      "Epoch 31/100\n",
      "44800/44800 [==============================] - 2s 47us/sample - loss: 0.5372 - accuracy: 0.7349 - val_loss: 0.5349 - val_accuracy: 0.7371\n",
      "Epoch 32/100\n",
      "44800/44800 [==============================] - 2s 44us/sample - loss: 0.5395 - accuracy: 0.7307 - val_loss: 0.5460 - val_accuracy: 0.7270\n",
      "Epoch 33/100\n",
      "44800/44800 [==============================] - 2s 43us/sample - loss: 0.5377 - accuracy: 0.7333 - val_loss: 0.5371 - val_accuracy: 0.7353\n",
      "Epoch 34/100\n",
      "44800/44800 [==============================] - 2s 47us/sample - loss: 0.5314 - accuracy: 0.7390 - val_loss: 0.5385 - val_accuracy: 0.7350- loss: 0.5321 - accu - ETA: 0s - loss: 0.5319 - accu\n",
      "Epoch 35/100\n",
      "44800/44800 [==============================] - 1s 32us/sample - loss: 0.5334 - accuracy: 0.7375 - val_loss: 0.5611 - val_accuracy: 0.7151\n",
      "Epoch 36/100\n",
      "44800/44800 [==============================] - 2s 40us/sample - loss: 0.5289 - accuracy: 0.7409 - val_loss: 0.5464 - val_accuracy: 0.7258\n",
      "Epoch 37/100\n",
      "44800/44800 [==============================] - 2s 34us/sample - loss: 0.5303 - accuracy: 0.7387 - val_loss: 0.5388 - val_accuracy: 0.7348\n",
      "Epoch 38/100\n",
      "44800/44800 [==============================] - 2s 40us/sample - loss: 0.5380 - accuracy: 0.7344 - val_loss: 0.5356 - val_accuracy: 0.7358\n",
      "Epoch 39/100\n",
      "44800/44800 [==============================] - 2s 37us/sample - loss: 0.5292 - accuracy: 0.7417 - val_loss: 0.5391 - val_accuracy: 0.7342\n",
      "Epoch 40/100\n",
      "44800/44800 [==============================] - 2s 39us/sample - loss: 0.5361 - accuracy: 0.7349 - val_loss: 0.5500 - val_accuracy: 0.7219\n",
      "Epoch 41/100\n",
      "44800/44800 [==============================] - 2s 44us/sample - loss: 0.5372 - accuracy: 0.7336 - val_loss: 0.5521 - val_accuracy: 0.7253\n",
      "Epoch 42/100\n",
      "44800/44800 [==============================] - 2s 51us/sample - loss: 0.5315 - accuracy: 0.7369 - val_loss: 0.5427 - val_accuracy: 0.7307\n",
      "Epoch 43/100\n",
      "44800/44800 [==============================] - 3s 58us/sample - loss: 0.5350 - accuracy: 0.7350 - val_loss: 0.5418 - val_accuracy: 0.7321\n",
      "Epoch 44/100\n",
      "44800/44800 [==============================] - 2s 52us/sample - loss: 0.5332 - accuracy: 0.7358 - val_loss: 0.5364 - val_accuracy: 0.7388\n",
      "Epoch 45/100\n",
      "44800/44800 [==============================] - 2s 50us/sample - loss: 0.5375 - accuracy: 0.7340 - val_loss: 0.5408 - val_accuracy: 0.7391\n",
      "Epoch 46/100\n",
      "44800/44800 [==============================] - 2s 37us/sample - loss: 0.5353 - accuracy: 0.7355 - val_loss: 0.5362 - val_accuracy: 0.7368\n",
      "Epoch 47/100\n",
      "44800/44800 [==============================] - 2s 36us/sample - loss: 0.5387 - accuracy: 0.7326 - val_loss: 0.5370 - val_accuracy: 0.7367\n",
      "Epoch 48/100\n",
      "44800/44800 [==============================] - 2s 36us/sample - loss: 0.5302 - accuracy: 0.7381 - val_loss: 0.5374 - val_accuracy: 0.7358\n",
      "Epoch 49/100\n",
      "44800/44800 [==============================] - 2s 49us/sample - loss: 0.5309 - accuracy: 0.7398 - val_loss: 0.5634 - val_accuracy: 0.7188\n",
      "Epoch 50/100\n",
      "44800/44800 [==============================] - 2s 44us/sample - loss: 0.5343 - accuracy: 0.7362 - val_loss: 0.5409 - val_accuracy: 0.7330\n",
      "Epoch 51/100\n",
      "44800/44800 [==============================] - 3s 59us/sample - loss: 0.5312 - accuracy: 0.7387 - val_loss: 0.5422 - val_accuracy: 0.7319\n",
      "Epoch 52/100\n",
      "44800/44800 [==============================] - 1s 33us/sample - loss: 0.5314 - accuracy: 0.7372 - val_loss: 0.5369 - val_accuracy: 0.7341\n",
      "Epoch 53/100\n",
      "44800/44800 [==============================] - 2s 41us/sample - loss: 0.5340 - accuracy: 0.7374 - val_loss: 0.5457 - val_accuracy: 0.7299s - loss: 0.5322 - ac\n",
      "Epoch 54/100\n",
      "44800/44800 [==============================] - 2s 49us/sample - loss: 0.5333 - accuracy: 0.7376 - val_loss: 0.5369 - val_accuracy: 0.7360\n",
      "Epoch 55/100\n",
      "44800/44800 [==============================] - 2s 38us/sample - loss: 0.5309 - accuracy: 0.7394 - val_loss: 0.5423 - val_accuracy: 0.7307\n",
      "Epoch 56/100\n",
      "44800/44800 [==============================] - 2s 39us/sample - loss: 0.5293 - accuracy: 0.7412 - val_loss: 0.5372 - val_accuracy: 0.7362\n",
      "Epoch 57/100\n",
      "44800/44800 [==============================] - 2s 36us/sample - loss: 0.5329 - accuracy: 0.7374 - val_loss: 0.5383 - val_accuracy: 0.7379\n",
      "Epoch 58/100\n",
      "44800/44800 [==============================] - 2s 49us/sample - loss: 0.5312 - accuracy: 0.7375 - val_loss: 0.5414 - val_accuracy: 0.7352\n",
      "Epoch 59/100\n",
      "44800/44800 [==============================] - 2s 35us/sample - loss: 0.5338 - accuracy: 0.7362 - val_loss: 0.5414 - val_accuracy: 0.7336\n",
      "Epoch 60/100\n",
      "44800/44800 [==============================] - 2s 40us/sample - loss: 0.5321 - accuracy: 0.7375 - val_loss: 0.5384 - val_accuracy: 0.7379s - loss: 0.5320 - accura\n",
      "Epoch 61/100\n",
      "44800/44800 [==============================] - 2s 36us/sample - loss: 0.5303 - accuracy: 0.7370 - val_loss: 0.5382 - val_accuracy: 0.738008 - accuracy: \n",
      "Epoch 62/100\n",
      "44800/44800 [==============================] - 1s 33us/sample - loss: 0.5294 - accuracy: 0.7391 - val_loss: 0.5526 - val_accuracy: 0.7296\n",
      "Epoch 63/100\n",
      "44800/44800 [==============================] - 2s 36us/sample - loss: 0.5359 - accuracy: 0.7344 - val_loss: 0.5477 - val_accuracy: 0.7267\n",
      "Epoch 64/100\n",
      "44800/44800 [==============================] - 2s 37us/sample - loss: 0.5356 - accuracy: 0.7346 - val_loss: 0.5487 - val_accuracy: 0.7307\n",
      "Epoch 65/100\n",
      "44800/44800 [==============================] - 2s 37us/sample - loss: 0.5293 - accuracy: 0.7396 - val_loss: 0.5409 - val_accuracy: 0.7344\n",
      "Epoch 66/100\n",
      "44800/44800 [==============================] - 2s 39us/sample - loss: 0.5351 - accuracy: 0.7337 - val_loss: 0.5455 - val_accuracy: 0.7317\n",
      "Epoch 67/100\n",
      "44800/44800 [==============================] - 2s 51us/sample - loss: 0.5357 - accuracy: 0.7353 - val_loss: 0.5459 - val_accuracy: 0.7294\n",
      "Epoch 68/100\n",
      "44800/44800 [==============================] - 2s 41us/sample - loss: 0.5353 - accuracy: 0.7340 - val_loss: 0.5688 - val_accuracy: 0.7121\n",
      "Epoch 69/100\n",
      "44800/44800 [==============================] - 2s 40us/sample - loss: 0.5365 - accuracy: 0.7336 - val_loss: 0.6113 - val_accuracy: 0.6510\n",
      "Epoch 70/100\n",
      "44800/44800 [==============================] - 2s 45us/sample - loss: 0.5400 - accuracy: 0.7306 - val_loss: 0.5470 - val_accuracy: 0.7256\n",
      "Epoch 71/100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "44800/44800 [==============================] - 2s 39us/sample - loss: 0.5317 - accuracy: 0.7363 - val_loss: 0.5456 - val_accuracy: 0.7312\n",
      "Epoch 72/100\n",
      "44800/44800 [==============================] - 1s 32us/sample - loss: 0.5324 - accuracy: 0.7365 - val_loss: 0.5399 - val_accuracy: 0.7341\n",
      "Epoch 73/100\n",
      "44800/44800 [==============================] - 2s 41us/sample - loss: 0.5277 - accuracy: 0.7412 - val_loss: 0.5487 - val_accuracy: 0.7297\n",
      "Epoch 74/100\n",
      "44800/44800 [==============================] - 1s 30us/sample - loss: 0.5331 - accuracy: 0.7371 - val_loss: 0.5472 - val_accuracy: 0.7276\n",
      "Epoch 75/100\n",
      "44800/44800 [==============================] - 2s 48us/sample - loss: 0.5394 - accuracy: 0.7304 - val_loss: 0.5754 - val_accuracy: 0.6996\n",
      "Epoch 76/100\n",
      "44800/44800 [==============================] - 2s 45us/sample - loss: 0.5328 - accuracy: 0.7360 - val_loss: 0.5484 - val_accuracy: 0.7275\n",
      "Epoch 77/100\n",
      "44800/44800 [==============================] - 2s 51us/sample - loss: 0.5286 - accuracy: 0.7409 - val_loss: 0.5432 - val_accuracy: 0.7306\n",
      "Epoch 78/100\n",
      "44800/44800 [==============================] - 3s 56us/sample - loss: 0.5298 - accuracy: 0.7403 - val_loss: 0.5389 - val_accuracy: 0.7359\n",
      "Epoch 79/100\n",
      "44800/44800 [==============================] - 2s 41us/sample - loss: 0.5281 - accuracy: 0.7397 - val_loss: 0.5398 - val_accuracy: 0.7340\n",
      "Epoch 80/100\n",
      "44800/44800 [==============================] - 2s 38us/sample - loss: 0.5350 - accuracy: 0.7346 - val_loss: 0.5417 - val_accuracy: 0.7351\n",
      "Epoch 81/100\n",
      "44800/44800 [==============================] - 2s 34us/sample - loss: 0.5276 - accuracy: 0.7404 - val_loss: 0.5442 - val_accuracy: 0.7334-\n",
      "Epoch 82/100\n",
      "44800/44800 [==============================] - 2s 39us/sample - loss: 0.5282 - accuracy: 0.7409 - val_loss: 0.5405 - val_accuracy: 0.7362 - \n",
      "Epoch 83/100\n",
      "44800/44800 [==============================] - 2s 40us/sample - loss: 0.5289 - accuracy: 0.7418 - val_loss: 0.5592 - val_accuracy: 0.7153\n",
      "Epoch 84/100\n",
      "44800/44800 [==============================] - 2s 41us/sample - loss: 0.5312 - accuracy: 0.7388 - val_loss: 0.5386 - val_accuracy: 0.7354\n",
      "Epoch 85/100\n",
      "44800/44800 [==============================] - 2s 48us/sample - loss: 0.5357 - accuracy: 0.7343 - val_loss: 0.5439 - val_accuracy: 0.7321\n",
      "Epoch 86/100\n",
      "44800/44800 [==============================] - 2s 44us/sample - loss: 0.5281 - accuracy: 0.7400 - val_loss: 0.5399 - val_accuracy: 0.7361ss: 0.5292 - \n",
      "Epoch 87/100\n",
      "44800/44800 [==============================] - 2s 46us/sample - loss: 0.5303 - accuracy: 0.7385 - val_loss: 0.5415 - val_accuracy: 0.7361\n",
      "Epoch 88/100\n",
      "44800/44800 [==============================] - 2s 45us/sample - loss: 0.5300 - accuracy: 0.7387 - val_loss: 0.5401 - val_accuracy: 0.7364\n",
      "Epoch 89/100\n",
      "44800/44800 [==============================] - 2s 38us/sample - loss: 0.5276 - accuracy: 0.7406 - val_loss: 0.5411 - val_accuracy: 0.7342\n",
      "Epoch 90/100\n",
      "44800/44800 [==============================] - 2s 45us/sample - loss: 0.5304 - accuracy: 0.7379 - val_loss: 0.5464 - val_accuracy: 0.7276\n",
      "Epoch 91/100\n",
      "44800/44800 [==============================] - 2s 50us/sample - loss: 0.5269 - accuracy: 0.7407 - val_loss: 0.5582 - val_accuracy: 0.7229\n",
      "Epoch 92/100\n",
      "44800/44800 [==============================] - 2s 44us/sample - loss: 0.5332 - accuracy: 0.7361 - val_loss: 0.5416 - val_accuracy: 0.7347\n",
      "Epoch 93/100\n",
      "44800/44800 [==============================] - 2s 47us/sample - loss: 0.5300 - accuracy: 0.7390 - val_loss: 0.5422 - val_accuracy: 0.7337\n",
      "Epoch 94/100\n",
      "44800/44800 [==============================] - 2s 37us/sample - loss: 0.5268 - accuracy: 0.7422 - val_loss: 0.5680 - val_accuracy: 0.7090\n",
      "Epoch 95/100\n",
      "44800/44800 [==============================] - 2s 37us/sample - loss: 0.5301 - accuracy: 0.7373 - val_loss: 0.5501 - val_accuracy: 0.7248\n",
      "Epoch 96/100\n",
      "44800/44800 [==============================] - 2s 35us/sample - loss: 0.5287 - accuracy: 0.7399 - val_loss: 0.5424 - val_accuracy: 0.7341\n",
      "Epoch 97/100\n",
      "44800/44800 [==============================] - 1s 32us/sample - loss: 0.5304 - accuracy: 0.7363 - val_loss: 0.5529 - val_accuracy: 0.7247\n",
      "Epoch 98/100\n",
      "44800/44800 [==============================] - 1s 31us/sample - loss: 0.5310 - accuracy: 0.7389 - val_loss: 0.5627 - val_accuracy: 0.7176\n",
      "Epoch 99/100\n",
      "44800/44800 [==============================] - 1s 30us/sample - loss: 0.5319 - accuracy: 0.7358 - val_loss: 0.5399 - val_accuracy: 0.7340\n",
      "Epoch 100/100\n",
      "44800/44800 [==============================] - 1s 33us/sample - loss: 0.5365 - accuracy: 0.7335 - val_loss: 0.5400 - val_accuracy: 0.7361\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "    positive       0.72      0.78      0.75      5639\n",
      "    negative       0.76      0.69      0.72      5561\n",
      "\n",
      "    accuracy                           0.74     11200\n",
      "   macro avg       0.74      0.74      0.74     11200\n",
      "weighted avg       0.74      0.74      0.74     11200\n",
      "\n",
      "Accuracy  : 0.7360714285714286\n",
      "Precision : 0.738016693278231\n",
      "f1Score : 0.7353962320342924\n",
      "[[4425 1214]\n",
      " [1742 3819]]\n"
     ]
    }
   ],
   "source": [
    "#======== Cross Validation ===========\n",
    "skf = StratifiedKFold(n_splits=5, shuffle=True)\n",
    "skf.get_n_splits(xTrain, yTrain)\n",
    "foldNum=0\n",
    "for train_index, val_index in skf.split(xTrain, yTrain):\n",
    "    foldNum+=1\n",
    "    print(\"Results for fold\",foldNum)\n",
    "    X_train, X_val = X[train_index], X[val_index]\n",
    "    Y_train, Y_val = Y[train_index], Y[val_index]\n",
    "    \n",
    "    # one hot encode\n",
    "    Y_train = to_categorical(Y_train)\n",
    "    Y_val = to_categorical(Y_val)\n",
    "    \n",
    "    history = model.fit(X_train, Y_train, \n",
    "                        validation_data = (X_val, Y_val), \n",
    "                        epochs=epochs, \n",
    "                        batch_size=batch_size)  \n",
    "    yPredict = model.predict(X_val)\n",
    "\n",
    "    #Converting one hot encoded test label to label    \n",
    "    pred = np.argmax(yPredict, axis=1)\n",
    "    val = np.argmax(Y_val, axis=1)\n",
    "    \n",
    "    showResults(val, pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "    positive       0.72      0.79      0.75      6979\n",
      "    negative       0.77      0.69      0.73      7021\n",
      "\n",
      "    accuracy                           0.74     14000\n",
      "   macro avg       0.74      0.74      0.74     14000\n",
      "weighted avg       0.74      0.74      0.74     14000\n",
      "\n",
      "Accuracy  : 0.7383571428571428\n",
      "Precision : 0.7409640648250261\n",
      "f1Score : 0.7377095732106843\n",
      "[[5506 1473]\n",
      " [2190 4831]]\n"
     ]
    }
   ],
   "source": [
    "#============Test Phase============\n",
    "yPred = model.predict(xTest)\n",
    "yTest = to_categorical(yTest)\n",
    "pred = np.argmax(yPred, axis=1)\n",
    "test = np.argmax(yTest, axis=1)\n",
    "showResults(test, pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "#========== Save the trained mode for future use ==========\n",
    "model.save(\"trainedModel.h5\")"
   ]
  }
 ],
 "metadata": {
  "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": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}