5WEEK_REUTER_DATASET-Copy1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "ac69ecd7",
   "metadata": {},
   "source": [
    "# REUTER DATASET\n",
    "\n",
    "목표: 로이터 뉴스를 46개의 토픽으로 분류한다. \n",
    "cf.다중분류 문제의 예. 각 뉴스가 하나의 범주로 분류됨으로 단일 레이블 다중 분류다. 만일 여러개의 범주로 나뉜다면 그것은 다중 레이블 다중 분류 문제일 것이다.\n",
    "\n",
    "구성: 46개의 토픽으로 구성. 각 토픽은 10개 이상의 샘플로 이루어짐."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "fba5aca1",
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.datasets import reuters\n",
    "\n",
    "(train_data,train_labels), (test_data, test_labels) = reuters.load_data(num_words=10000)\n",
    "#IMDB처럼 num_words 매개변수는 데이터에서 가장 자주 등장하는 단어 1만개로 제한\n",
    "#train 데이터는 IMDB처럼 정수 리스트"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "86d6f1b5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "word_index = reuters.get_word_index()\n",
    "reverse_word_index = dict([(value, key) for (key, value) in word_index.items()] )\n",
    "decoded_newswire = ' '.join([reverse_word_index.get(i - 3 , '?') for i in train_data[0]]) \n",
    "#0,1,2는 패딩, 문서 시작, 사전에 없음을 위한 인덱스이므로 3을 뺀다 cf. IMDB도 동일\n",
    "train_labels[10] #샘플에 연결된 레이블은 토픽의 인덱스로 0~45 사이의 정수"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "8234e7e9",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "\n",
    "def vectorize_sequences(sequences, dimension=10000): #one-hot encoding함수 구현\n",
    "    results = np.zeros((len(sequences),dimension)) #행:sequences, 열:dimension 크기의 원소가 0인 배열 생성\n",
    "    for i, sequence in enumerate(sequences): #enumerate는 (리스트의 원소 인덱스, 원소)로 반환한다. -> i:인덱스 sequence:원소 \n",
    "        results[i,sequence] = 1. #results[i]에서 특정 인덱스의 위치를 1로 만든다\n",
    "    return results\n",
    "\n",
    "x_train = vectorize_sequences(train_data) #훈련 데이터 백터 변환\n",
    "x_test = vectorize_sequences(test_data) #테스트 데이터 백터 변환\n",
    "one_hot_train_labels = vectorize_sequences(train_labels,46) # 레이블은 0~45까지의 인덱스이므로 demention을 46으로 해야함\n",
    "one_hot_test_labels = vectorize_sequences(test_labels,46)\n",
    "\n",
    "#cf.MNIST예제처럼 to_categorical 내장함수를 사용해도 됨"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "22805601",
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras import models\n",
    "from keras import layers\n",
    "\n",
    "model = models.Sequential()\n",
    "model.add(layers.Dense(64,activation='relu',input_shape=(10000,))) #Output layer가 46개. 따라서 hidden layer는 이보다 많아야함\n",
    "model.add(layers.Dense(64,activation='relu'))\n",
    "model.add(layers.Dense(46,activation='softmax')) #MNIST와 동일하게 각 입력 샘플마다 46개의 출력 클래스에 대한 확률 분포를 출력\n",
    "                                                 #46개의 값을 더하면 1이다"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "c11a8077",
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(optimizer='rmsprop',\n",
    "             loss='categorical_crossentropy',\n",
    "             metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "67632916",
   "metadata": {},
   "outputs": [],
   "source": [
    "x_val = x_train[:1000] #처음부터 인덱스9999까지\n",
    "partial_x_train = x_train[1000:] #10000부터 끝까지\n",
    "y_val = one_hot_train_labels[:1000] #검증할 데이터\n",
    "partial_y_train = one_hot_train_labels[1000:] #훈련시킬 데이터"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "1415b3ba",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "16/16 [==============================] - 1s 31ms/step - loss: 2.5993 - accuracy: 0.5331 - val_loss: 1.7253 - val_accuracy: 0.6310\n",
      "Epoch 2/20\n",
      "16/16 [==============================] - 0s 16ms/step - loss: 1.4111 - accuracy: 0.6994 - val_loss: 1.2994 - val_accuracy: 0.7130\n",
      "Epoch 3/20\n",
      "16/16 [==============================] - 0s 17ms/step - loss: 1.0459 - accuracy: 0.7711 - val_loss: 1.1145 - val_accuracy: 0.7520\n",
      "Epoch 4/20\n",
      "16/16 [==============================] - 0s 17ms/step - loss: 0.8211 - accuracy: 0.8275 - val_loss: 1.0244 - val_accuracy: 0.7730\n",
      "Epoch 5/20\n",
      "16/16 [==============================] - 0s 22ms/step - loss: 0.6482 - accuracy: 0.8639 - val_loss: 0.9579 - val_accuracy: 0.7910\n",
      "Epoch 6/20\n",
      "16/16 [==============================] - 0s 19ms/step - loss: 0.5151 - accuracy: 0.8918 - val_loss: 0.9773 - val_accuracy: 0.7800\n",
      "Epoch 7/20\n",
      "16/16 [==============================] - 0s 18ms/step - loss: 0.4178 - accuracy: 0.9121 - val_loss: 0.8963 - val_accuracy: 0.8170\n",
      "Epoch 8/20\n",
      "16/16 [==============================] - 0s 18ms/step - loss: 0.3371 - accuracy: 0.9297 - val_loss: 0.9099 - val_accuracy: 0.8070\n",
      "Epoch 9/20\n",
      "16/16 [==============================] - 0s 18ms/step - loss: 0.2853 - accuracy: 0.9377 - val_loss: 0.8786 - val_accuracy: 0.8190\n",
      "Epoch 10/20\n",
      "16/16 [==============================] - 0s 17ms/step - loss: 0.2347 - accuracy: 0.9451 - val_loss: 0.8811 - val_accuracy: 0.8240\n",
      "Epoch 11/20\n",
      "16/16 [==============================] - 0s 17ms/step - loss: 0.2020 - accuracy: 0.9491 - val_loss: 0.9222 - val_accuracy: 0.8140\n",
      "Epoch 12/20\n",
      "16/16 [==============================] - 0s 24ms/step - loss: 0.1861 - accuracy: 0.9493 - val_loss: 0.9226 - val_accuracy: 0.8150\n",
      "Epoch 13/20\n",
      "16/16 [==============================] - 0s 18ms/step - loss: 0.1641 - accuracy: 0.9530 - val_loss: 0.9326 - val_accuracy: 0.8140\n",
      "Epoch 14/20\n",
      "16/16 [==============================] - 0s 17ms/step - loss: 0.1540 - accuracy: 0.9533 - val_loss: 0.9881 - val_accuracy: 0.8020\n",
      "Epoch 15/20\n",
      "16/16 [==============================] - 0s 17ms/step - loss: 0.1394 - accuracy: 0.9559 - val_loss: 1.0249 - val_accuracy: 0.8050\n",
      "Epoch 16/20\n",
      "16/16 [==============================] - 0s 22ms/step - loss: 0.1336 - accuracy: 0.9559 - val_loss: 1.0397 - val_accuracy: 0.8020\n",
      "Epoch 17/20\n",
      "16/16 [==============================] - 0s 18ms/step - loss: 0.1292 - accuracy: 0.9565 - val_loss: 1.0065 - val_accuracy: 0.8120\n",
      "Epoch 18/20\n",
      "16/16 [==============================] - 0s 19ms/step - loss: 0.1210 - accuracy: 0.9564 - val_loss: 1.0438 - val_accuracy: 0.8100\n",
      "Epoch 19/20\n",
      "16/16 [==============================] - 0s 18ms/step - loss: 0.1152 - accuracy: 0.9564 - val_loss: 1.0703 - val_accuracy: 0.8030\n",
      "Epoch 20/20\n",
      "16/16 [==============================] - 0s 16ms/step - loss: 0.1140 - accuracy: 0.9570 - val_loss: 1.1380 - val_accuracy: 0.7950\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(partial_x_train,\n",
    "                   partial_y_train,\n",
    "                   epochs=20,\n",
    "                   batch_size=512,\n",
    "                   validation_data=(x_val,y_val))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "5de93e39",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "history_dict = history.history\n",
    "loss_values = history_dict['loss'] \n",
    "val_loss_values = history_dict['val_loss']\n",
    "\n",
    "epochs = range(1, len(loss_values) + 1)\n",
    "\n",
    "plt.plot(epochs, loss_values, 'bo', label='Training loss') \n",
    "plt.plot(epochs, val_loss_values, 'b', label='Validation loss')\n",
    "plt.title('Training and validation loss') \n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Loss') \n",
    "plt.legend()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "f29b43db",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.clf() # 생성한 그래프 초기화\n",
    "acc = history_dict['accuracy'] \n",
    "val_acc = history_dict['val_accuracy']\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.xlabel('Epochs')\n",
    "plt.ylabel('Acc') \n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "c5033953",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/9\n",
      "16/16 [==============================] - 1s 18ms/step - loss: 0.3755 - accuracy: 0.0375\n",
      "Epoch 2/9\n",
      "16/16 [==============================] - 0s 19ms/step - loss: 0.1174 - accuracy: 0.4214\n",
      "Epoch 3/9\n",
      "16/16 [==============================] - 0s 18ms/step - loss: 0.0748 - accuracy: 0.5677\n",
      "Epoch 4/9\n",
      "16/16 [==============================] - 0s 17ms/step - loss: 0.0555 - accuracy: 0.6612\n",
      "Epoch 5/9\n",
      "16/16 [==============================] - 0s 18ms/step - loss: 0.0437 - accuracy: 0.7189\n",
      "Epoch 6/9\n",
      "16/16 [==============================] - 0s 19ms/step - loss: 0.0353 - accuracy: 0.7734\n",
      "Epoch 7/9\n",
      "16/16 [==============================] - 0s 17ms/step - loss: 0.0295 - accuracy: 0.8107\n",
      "Epoch 8/9\n",
      "16/16 [==============================] - 0s 20ms/step - loss: 0.0249 - accuracy: 0.8439\n",
      "Epoch 9/9\n",
      "16/16 [==============================] - 0s 18ms/step - loss: 0.0214 - accuracy: 0.8721\n",
      "71/71 [==============================] - 0s 2ms/step - loss: 0.0338 - accuracy: 0.7756\n",
      "[0.033801719546318054, 0.7756010890007019]\n"
     ]
    }
   ],
   "source": [
    "#9번째 epochs에서부터 overfitting되었음 \n",
    "#따라서 epochs를 9까지로 제한하고 훈련 재시작\n",
    "model = models.Sequential()\n",
    "model.add(layers.Dense(64, activation='relu', input_shape=(10000,))) \n",
    "model.add(layers.Dense(64, activation='relu')) \n",
    "model.add(layers.Dense(46, activation='sigmoid'))\n",
    "\n",
    "model.compile(optimizer='rmsprop',\n",
    "              loss='binary_crossentropy',\n",
    "              metrics=['accuracy'])\n",
    "\n",
    "model.fit(partial_x_train, partial_y_train, epochs=9, batch_size=512) \n",
    "results = model.evaluate(x_test, one_hot_test_labels)\n",
    "print(results)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}