6.2-understanding-recurrent-neural-networks.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "'2.2.2'"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import keras\n",
    "keras.__version__"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Understanding recurrent neural networks\n",
    "\n",
    "이 노트북은 [케라스 창시자에게 배우는 딥러닝](https://tensorflow.blog/%EC%BC%80%EB%9D%BC%EC%8A%A4-%EB%94%A5%EB%9F%AC%EB%8B%9D/) 책의 6장 2절의 코드 예제입니다. 책에는 더 많은 내용과 그림이 있습니다. 이 노트북에는 소스 코드에 관련된 설명만 포함합니다.\n",
    "\n",
    "---\n",
    "\n",
    "[...]\n",
    "\n",
    "## 케라스의 순환 층\n",
    "\n",
    "\n",
    "넘파이로 간단하게 구현한 과정이 실제 케라스의 `SimpleRNN` 층에 해당합니다:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.layers import SimpleRNN"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`SimpleRNN`이 한 가지 다른 점은 넘파이 예제처럼 하나의 시퀀스가 아니라 다른 케라스 층과 마찬가지로 시퀀스 배치를 처리한다는 것입니다. 즉, `(timesteps, input_features)` 크기가 아니라 `(batch_size, timesteps, input_features)` 크기의 입력을 받습니다.\n",
    "\n",
    "케라스에 있는 모든 순환 층과 동일하게 `SimpleRNN`은 두 가지 모드로 실행할 수 있습니다. 각 타임스텝의 출력을 모은 전체 시퀀스를 반환하거나(크기가 `(batch_size, timesteps, output_features)`인 3D 텐서), 입력 시퀀스에 대한 마지막 출력만 반환할 수 있습니다(크기가 `(batch_size, output_features)`인 2D 텐서). 이 모드는 객체를 생성할 때 `return_sequences` 매개변수로 선택할 수 있습니다. 예제를 살펴보죠:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "embedding_1 (Embedding)      (None, None, 32)          320000    \n",
      "_________________________________________________________________\n",
      "simple_rnn_1 (SimpleRNN)     (None, 32)                2080      \n",
      "=================================================================\n",
      "Total params: 322,080\n",
      "Trainable params: 322,080\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import Embedding, SimpleRNN\n",
    "\n",
    "model = Sequential()\n",
    "model.add(Embedding(10000, 32))\n",
    "model.add(SimpleRNN(32))\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "embedding_2 (Embedding)      (None, None, 32)          320000    \n",
      "_________________________________________________________________\n",
      "simple_rnn_2 (SimpleRNN)     (None, None, 32)          2080      \n",
      "=================================================================\n",
      "Total params: 322,080\n",
      "Trainable params: 322,080\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model = Sequential()\n",
    "model.add(Embedding(10000, 32))\n",
    "model.add(SimpleRNN(32, return_sequences=True))\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "네트워크의 표현력을 증가시키기 위해 여러 개의 순환 층을 차례대로 쌓는 것이 유용할 때가 있습니다. 이런 설정에서는 중간 층들이 전체 출력 시퀀스를 반환하도록 설정해야 합니다:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "embedding_3 (Embedding)      (None, None, 32)          320000    \n",
      "_________________________________________________________________\n",
      "simple_rnn_3 (SimpleRNN)     (None, None, 32)          2080      \n",
      "_________________________________________________________________\n",
      "simple_rnn_4 (SimpleRNN)     (None, None, 32)          2080      \n",
      "_________________________________________________________________\n",
      "simple_rnn_5 (SimpleRNN)     (None, None, 32)          2080      \n",
      "_________________________________________________________________\n",
      "simple_rnn_6 (SimpleRNN)     (None, 32)                2080      \n",
      "=================================================================\n",
      "Total params: 328,320\n",
      "Trainable params: 328,320\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model = Sequential()\n",
    "model.add(Embedding(10000, 32))\n",
    "model.add(SimpleRNN(32, return_sequences=True))\n",
    "model.add(SimpleRNN(32, return_sequences=True))\n",
    "model.add(SimpleRNN(32, return_sequences=True))\n",
    "model.add(SimpleRNN(32))  # 맨 위 층만 마지막 출력을 반환합니다.\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "이제 IMDB 영화 리뷰 분류 문제에 적용해 보죠. 먼저 데이터를 전처리합니다:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "데이터 로딩...\n",
      "25000 훈련 시퀀스\n",
      "25000 테스트 시퀀스\n",
      "시퀀스 패딩 (samples x time)\n",
      "input_train 크기: (25000, 500)\n",
      "input_test 크기: (25000, 500)\n"
     ]
    }
   ],
   "source": [
    "from keras.datasets import imdb\n",
    "from keras.preprocessing import sequence\n",
    "\n",
    "max_features = 10000  # 특성으로 사용할 단어의 수\n",
    "maxlen = 500  # 사용할 텍스트의 길이(가장 빈번한 max_features 개의 단어만 사용합니다)\n",
    "batch_size = 32\n",
    "\n",
    "print('데이터 로딩...')\n",
    "(input_train, y_train), (input_test, y_test) = imdb.load_data(num_words=max_features)\n",
    "print(len(input_train), '훈련 시퀀스')\n",
    "print(len(input_test), '테스트 시퀀스')\n",
    "\n",
    "print('시퀀스 패딩 (samples x time)')\n",
    "input_train = sequence.pad_sequences(input_train, maxlen=maxlen)\n",
    "input_test = sequence.pad_sequences(input_test, maxlen=maxlen)\n",
    "print('input_train 크기:', input_train.shape)\n",
    "print('input_test 크기:', input_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`Embedding` 층과 `SimpleRNN` 층을 사용해 간단한 순환 네트워크를 훈련시켜 보겠습니다:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 20000 samples, validate on 5000 samples\n",
      "Epoch 1/10\n",
      "20000/20000 [==============================] - 16s 806us/step - loss: 0.6420 - acc: 0.6135 - val_loss: 0.5809 - val_acc: 0.6910\n",
      "Epoch 2/10\n",
      "20000/20000 [==============================] - 15s 774us/step - loss: 0.4127 - acc: 0.8219 - val_loss: 0.3867 - val_acc: 0.8366\n",
      "Epoch 3/10\n",
      "20000/20000 [==============================] - 15s 774us/step - loss: 0.2981 - acc: 0.8813 - val_loss: 0.4465 - val_acc: 0.7854\n",
      "Epoch 4/10\n",
      "20000/20000 [==============================] - 15s 773us/step - loss: 0.2324 - acc: 0.9131 - val_loss: 0.3941 - val_acc: 0.8312\n",
      "Epoch 5/10\n",
      "20000/20000 [==============================] - 15s 773us/step - loss: 0.1772 - acc: 0.9343 - val_loss: 0.3593 - val_acc: 0.8740\n",
      "Epoch 6/10\n",
      "20000/20000 [==============================] - 15s 772us/step - loss: 0.1259 - acc: 0.9553 - val_loss: 0.3869 - val_acc: 0.8670\n",
      "Epoch 7/10\n",
      "20000/20000 [==============================] - 15s 771us/step - loss: 0.0789 - acc: 0.9746 - val_loss: 0.4566 - val_acc: 0.8508\n",
      "Epoch 8/10\n",
      "20000/20000 [==============================] - 15s 771us/step - loss: 0.0488 - acc: 0.9850 - val_loss: 0.6158 - val_acc: 0.8096\n",
      "Epoch 9/10\n",
      "20000/20000 [==============================] - 15s 770us/step - loss: 0.0358 - acc: 0.9899 - val_loss: 0.5539 - val_acc: 0.8384\n",
      "Epoch 10/10\n",
      "20000/20000 [==============================] - 15s 771us/step - loss: 0.0198 - acc: 0.9946 - val_loss: 0.6967 - val_acc: 0.7902\n"
     ]
    }
   ],
   "source": [
    "from keras.layers import Dense\n",
    "\n",
    "model = Sequential()\n",
    "model.add(Embedding(max_features, 32))\n",
    "model.add(SimpleRNN(32))\n",
    "model.add(Dense(1, activation='sigmoid'))\n",
    "\n",
    "model.compile(optimizer='rmsprop', loss='binary_crossentropy', metrics=['acc'])\n",
    "history = model.fit(input_train, y_train,\n",
    "                    epochs=10,\n",
    "                    batch_size=128,\n",
    "                    validation_split=0.2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "이제 훈련과 검증의 손실과 정확도를 그래프로 그립니다:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "acc = history.history['acc']\n",
    "val_acc = history.history['val_acc']\n",
    "loss = history.history['loss']\n",
    "val_loss = history.history['val_loss']\n",
    "\n",
    "epochs = range(1, len(acc) + 1)\n",
    "\n",
    "plt.plot(epochs, acc, 'bo', label='Training acc')\n",
    "plt.plot(epochs, val_acc, 'b', label='Validation acc')\n",
    "plt.title('Training and validation accuracy')\n",
    "plt.legend()\n",
    "\n",
    "plt.figure()\n",
    "\n",
    "plt.plot(epochs, loss, 'bo', label='Training loss')\n",
    "plt.plot(epochs, val_loss, 'b', label='Validation loss')\n",
    "plt.title('Training and validation loss')\n",
    "plt.legend()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "3장에서 이 데이터셋을 사용한 첫 번째 모델에서 얻은 테스트 정확도는 87%였습니다. 안타깝지만 간단한 순환 네트워크는 이 기준 모델보다 성능이 높지 않습니다(85% 정도의 검증 정확도를 얻었습니다). 이런 원인은 전체 시퀀스가 아니라 처음 500개의 단어만 입력에 사용했기 때문입니다. 이 RNN은 기준 모델보다 얻은 정보가 적습니다. 다른 이유는 `SimpleRNN`이 텍스트와 같이 긴 시퀀스를 처리하는데 적합하지 않기 때문입니다. 더 잘 작동하는 다른 순환 층이 있습니다. 조금 더 고급 순환 층을 살펴보죠."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[...]\n",
    "\n",
    "## 케라스를 사용한 LSTM 예제\n",
    "\n",
    "이제 실제적인 관심사로 이동해 보죠. LSTM 층으로 모델을 구성하고 IMDB 데이터에서 훈련해 보겠습니다(그림 6-16과 6-17 참조). 이 네트워크는 조금 전 `SimpleRNN`을 사용했던 모델과 비슷합니다. LSTM 층은 출력 차원만 지정하고 다른 (많은) 매개변수는 케라스의 기본값으로 남겨 두었습니다. 케라스는 좋은 기본값을 가지고 있어서 직접 매개변수를 튜닝하는 데 시간을 쓰지 않고도 거의 항상 어느정도 작동하는 모델을 얻을 수 있습니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 20000 samples, validate on 5000 samples\n",
      "Epoch 1/10\n",
      "20000/20000 [==============================] - 72s 4ms/step - loss: 0.5064 - acc: 0.7643 - val_loss: 0.4322 - val_acc: 0.8072\n",
      "Epoch 2/10\n",
      "20000/20000 [==============================] - 72s 4ms/step - loss: 0.2899 - acc: 0.8860 - val_loss: 0.2752 - val_acc: 0.8884\n",
      "Epoch 3/10\n",
      "20000/20000 [==============================] - 72s 4ms/step - loss: 0.2342 - acc: 0.9099 - val_loss: 0.3421 - val_acc: 0.8756\n",
      "Epoch 4/10\n",
      "20000/20000 [==============================] - 72s 4ms/step - loss: 0.2000 - acc: 0.9255 - val_loss: 0.4002 - val_acc: 0.8666\n",
      "Epoch 5/10\n",
      "20000/20000 [==============================] - 72s 4ms/step - loss: 0.1741 - acc: 0.9356 - val_loss: 0.3663 - val_acc: 0.8816\n",
      "Epoch 6/10\n",
      "20000/20000 [==============================] - 72s 4ms/step - loss: 0.1539 - acc: 0.9432 - val_loss: 0.3408 - val_acc: 0.8766\n",
      "Epoch 7/10\n",
      "20000/20000 [==============================] - 72s 4ms/step - loss: 0.1441 - acc: 0.9510 - val_loss: 0.4799 - val_acc: 0.8650\n",
      "Epoch 8/10\n",
      "20000/20000 [==============================] - 72s 4ms/step - loss: 0.1294 - acc: 0.9541 - val_loss: 0.3486 - val_acc: 0.8716\n",
      "Epoch 9/10\n",
      "20000/20000 [==============================] - 72s 4ms/step - loss: 0.1177 - acc: 0.9588 - val_loss: 0.3399 - val_acc: 0.8880\n",
      "Epoch 10/10\n",
      "20000/20000 [==============================] - 72s 4ms/step - loss: 0.1069 - acc: 0.9634 - val_loss: 0.3843 - val_acc: 0.8836\n"
     ]
    }
   ],
   "source": [
    "from keras.layers import LSTM\n",
    "\n",
    "model = Sequential()\n",
    "model.add(Embedding(max_features, 32))\n",
    "model.add(LSTM(32))\n",
    "model.add(Dense(1, activation='sigmoid'))\n",
    "\n",
    "model.compile(optimizer='rmsprop',\n",
    "              loss='binary_crossentropy',\n",
    "              metrics=['acc'])\n",
    "history = model.fit(input_train, y_train,\n",
    "                    epochs=10,\n",
    "                    batch_size=128,\n",
    "                    validation_split=0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "acc = history.history['acc']\n",
    "val_acc = history.history['val_acc']\n",
    "loss = history.history['loss']\n",
    "val_loss = history.history['val_loss']\n",
    "\n",
    "epochs = range(1, len(acc) + 1)\n",
    "\n",
    "plt.plot(epochs, acc, 'bo', label='Training acc')\n",
    "plt.plot(epochs, val_acc, 'b', label='Validation acc')\n",
    "plt.title('Training and validation accuracy')\n",
    "plt.legend()\n",
    "\n",
    "plt.figure()\n",
    "\n",
    "plt.plot(epochs, loss, 'bo', label='Training loss')\n",
    "plt.plot(epochs, val_loss, 'b', label='Validation loss')\n",
    "plt.title('Training and validation loss')\n",
    "plt.legend()\n",
    "\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}