nb_ex2_2_ann_rg.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.3 시계열 데이터를 예측하는 회귀 ANN 구현\n",
    "보스턴 집값을 예측하는 사례로 회귀 ANN 구현 사례를 다룹니다.\n",
    "\n",
    "### 2.3.1 회귀 ANN 모델링\n",
    "1. 회귀 ANN 모델을 구현합니다. \n",
    "- 모델 구현을 위해 케라스 패키지에 들어 있는 서브패키지 layers와 models를 불러옵니다. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras import layers, models"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- 클래스를 만들고 클래스 생성자 함수에 사용될 신경망 계층을 정의합니다. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "class ANN(models.Model):\n",
    "    def __init__(self, Nin, Nh, Nout):\n",
    "        # Prepare network layers and activate functions\n",
    "        hidden = layers.Dense(Nh)\n",
    "        output = layers.Dense(Nout)\n",
    "        relu = layers.Activation('relu')\n",
    "\n",
    "        # Connect network elements\n",
    "        x = layers.Input(shape=(Nin,))\n",
    "        h = relu(hidden(x))\n",
    "        y = output(h)\n",
    "\n",
    "        super().__init__(x, y)\n",
    "        self.compile(loss='mse', optimizer='sgd')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.3.2 학습과 평가용 데이터 불러오기\n",
    "2. 학습과 평가에 사용할 Boston housing 데이터 셋을 불러오고 정규화를 진행합니다. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras import datasets\n",
    "from sklearn import preprocessing\n",
    "\n",
    "def Data_func():\n",
    "    (X_train, y_train), (X_test, y_test) = datasets.boston_housing.load_data()\n",
    "    scaler = preprocessing.MinMaxScaler()\n",
    "    X_train = scaler.fit_transform(X_train)\n",
    "    X_test = scaler.transform(X_test)\n",
    "    return (X_train, y_train), (X_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.3.3 회귀 ANN 학습 결과 그래프 구현\n",
    "3. 회귀 ANN의 학습 결과를 표시하는 그래프"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keraspp.skeras import plot_loss\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.3.4 회귀 ANN 학습 및 성능 분석\n",
    "4. 회귀 ANN에 대한 학습과 성능을 분석을 진행하는 명령을 내립니다. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "4/4 - 0s - loss: 478.3341 - val_loss: 106.3890\n",
      "Epoch 2/100\n",
      "4/4 - 0s - loss: 96.1139 - val_loss: 88.2056\n",
      "Epoch 3/100\n",
      "4/4 - 0s - loss: 77.3092 - val_loss: 127.1817\n",
      "Epoch 4/100\n",
      "4/4 - 0s - loss: 80.6153 - val_loss: 58.0730\n",
      "Epoch 5/100\n",
      "4/4 - 0s - loss: 52.8222 - val_loss: 58.5521\n",
      "Epoch 6/100\n",
      "4/4 - 0s - loss: 52.4533 - val_loss: 51.9659\n",
      "Epoch 7/100\n",
      "4/4 - 0s - loss: 46.3898 - val_loss: 46.3008\n",
      "Epoch 8/100\n",
      "4/4 - 0s - loss: 44.0762 - val_loss: 60.0542\n",
      "Epoch 9/100\n",
      "4/4 - 0s - loss: 46.6453 - val_loss: 64.2309\n",
      "Epoch 10/100\n",
      "4/4 - 0s - loss: 54.2866 - val_loss: 46.7775\n",
      "Epoch 11/100\n",
      "4/4 - 0s - loss: 45.2620 - val_loss: 38.8658\n",
      "Epoch 12/100\n",
      "4/4 - 0s - loss: 43.7960 - val_loss: 60.6970\n",
      "Epoch 13/100\n",
      "4/4 - 0s - loss: 80.2178 - val_loss: 53.1720\n",
      "Epoch 14/100\n",
      "4/4 - 0s - loss: 57.6628 - val_loss: 39.2651\n",
      "Epoch 15/100\n",
      "4/4 - 0s - loss: 36.9601 - val_loss: 39.3463\n",
      "Epoch 16/100\n",
      "4/4 - 0s - loss: 52.3025 - val_loss: 91.1897\n",
      "Epoch 17/100\n",
      "4/4 - 0s - loss: 95.2121 - val_loss: 33.1031\n",
      "Epoch 18/100\n",
      "4/4 - 0s - loss: 42.7783 - val_loss: 35.0120\n",
      "Epoch 19/100\n",
      "4/4 - 0s - loss: 34.5063 - val_loss: 39.2311\n",
      "Epoch 20/100\n",
      "4/4 - 0s - loss: 32.6984 - val_loss: 33.8349\n",
      "Epoch 21/100\n",
      "4/4 - 0s - loss: 34.8321 - val_loss: 36.5389\n",
      "Epoch 22/100\n",
      "4/4 - 0s - loss: 39.4335 - val_loss: 29.7446\n",
      "Epoch 23/100\n",
      "4/4 - 0s - loss: 42.3343 - val_loss: 37.5604\n",
      "Epoch 24/100\n",
      "4/4 - 0s - loss: 36.6038 - val_loss: 38.3593\n",
      "Epoch 25/100\n",
      "4/4 - 0s - loss: 46.5404 - val_loss: 28.0824\n",
      "Epoch 26/100\n",
      "4/4 - 0s - loss: 28.4903 - val_loss: 41.7288\n",
      "Epoch 27/100\n",
      "4/4 - 0s - loss: 56.5827 - val_loss: 45.6351\n",
      "Epoch 28/100\n",
      "4/4 - 0s - loss: 48.5997 - val_loss: 27.0860\n",
      "Epoch 29/100\n",
      "4/4 - 0s - loss: 27.5989 - val_loss: 33.3803\n",
      "Epoch 30/100\n",
      "4/4 - 0s - loss: 37.8638 - val_loss: 25.8301\n",
      "Epoch 31/100\n",
      "4/4 - 0s - loss: 27.4941 - val_loss: 28.2079\n",
      "Epoch 32/100\n",
      "4/4 - 0s - loss: 33.9488 - val_loss: 56.1231\n",
      "Epoch 33/100\n",
      "4/4 - 0s - loss: 53.1116 - val_loss: 27.3879\n",
      "Epoch 34/100\n",
      "4/4 - 0s - loss: 26.9650 - val_loss: 31.0741\n",
      "Epoch 35/100\n",
      "4/4 - 0s - loss: 25.9162 - val_loss: 28.4108\n",
      "Epoch 36/100\n",
      "4/4 - 0s - loss: 24.9458 - val_loss: 33.1777\n",
      "Epoch 37/100\n",
      "4/4 - 0s - loss: 33.7092 - val_loss: 36.8526\n",
      "Epoch 38/100\n",
      "4/4 - 0s - loss: 29.2253 - val_loss: 49.2569\n",
      "Epoch 39/100\n",
      "4/4 - 0s - loss: 35.7495 - val_loss: 29.1518\n",
      "Epoch 40/100\n",
      "4/4 - 0s - loss: 24.1513 - val_loss: 27.0932\n",
      "Epoch 41/100\n",
      "4/4 - 0s - loss: 27.8277 - val_loss: 30.5038\n",
      "Epoch 42/100\n",
      "4/4 - 0s - loss: 24.6245 - val_loss: 31.2033\n",
      "Epoch 43/100\n",
      "4/4 - 0s - loss: 44.2243 - val_loss: 28.4233\n",
      "Epoch 44/100\n",
      "4/4 - 0s - loss: 29.5751 - val_loss: 30.2363\n",
      "Epoch 45/100\n",
      "4/4 - 0s - loss: 34.1515 - val_loss: 23.6316\n",
      "Epoch 46/100\n",
      "4/4 - 0s - loss: 29.2734 - val_loss: 21.9705\n",
      "Epoch 47/100\n",
      "4/4 - 0s - loss: 23.5655 - val_loss: 26.4632\n",
      "Epoch 48/100\n",
      "4/4 - 0s - loss: 29.6594 - val_loss: 21.5495\n",
      "Epoch 49/100\n",
      "4/4 - 0s - loss: 23.3636 - val_loss: 21.7324\n",
      "Epoch 50/100\n",
      "4/4 - 0s - loss: 23.7941 - val_loss: 25.7674\n",
      "Epoch 51/100\n",
      "4/4 - 0s - loss: 24.9940 - val_loss: 28.6282\n",
      "Epoch 52/100\n",
      "4/4 - 0s - loss: 24.9929 - val_loss: 27.3902\n",
      "Epoch 53/100\n",
      "4/4 - 0s - loss: 23.7171 - val_loss: 21.3664\n",
      "Epoch 54/100\n",
      "4/4 - 0s - loss: 22.8728 - val_loss: 27.5805\n",
      "Epoch 55/100\n",
      "4/4 - 0s - loss: 25.8092 - val_loss: 21.4810\n",
      "Epoch 56/100\n",
      "4/4 - 0s - loss: 21.8477 - val_loss: 24.3677\n",
      "Epoch 57/100\n",
      "4/4 - 0s - loss: 27.5235 - val_loss: 20.6347\n",
      "Epoch 58/100\n",
      "4/4 - 0s - loss: 23.6566 - val_loss: 22.2848\n",
      "Epoch 59/100\n",
      "4/4 - 0s - loss: 22.5610 - val_loss: 37.0005\n",
      "Epoch 60/100\n",
      "4/4 - 0s - loss: 32.9143 - val_loss: 19.6208\n",
      "Epoch 61/100\n",
      "4/4 - 0s - loss: 22.1534 - val_loss: 19.5799\n",
      "Epoch 62/100\n",
      "4/4 - 0s - loss: 23.3413 - val_loss: 41.1664\n",
      "Epoch 63/100\n",
      "4/4 - 0s - loss: 41.3568 - val_loss: 19.4826\n",
      "Epoch 64/100\n",
      "4/4 - 0s - loss: 23.1864 - val_loss: 22.7518\n",
      "Epoch 65/100\n",
      "4/4 - 0s - loss: 26.1769 - val_loss: 19.4488\n",
      "Epoch 66/100\n",
      "4/4 - 0s - loss: 28.2492 - val_loss: 39.5751\n",
      "Epoch 67/100\n",
      "4/4 - 0s - loss: 39.7458 - val_loss: 28.5972\n",
      "Epoch 68/100\n",
      "4/4 - 0s - loss: 30.7939 - val_loss: 21.5134\n",
      "Epoch 69/100\n",
      "4/4 - 0s - loss: 21.6307 - val_loss: 18.6919\n",
      "Epoch 70/100\n",
      "4/4 - 0s - loss: 21.8062 - val_loss: 18.6612\n",
      "Epoch 71/100\n",
      "4/4 - 0s - loss: 24.7820 - val_loss: 33.0950\n",
      "Epoch 72/100\n",
      "4/4 - 0s - loss: 39.6682 - val_loss: 29.2252\n",
      "Epoch 73/100\n",
      "4/4 - 0s - loss: 28.9979 - val_loss: 18.5206\n",
      "Epoch 74/100\n",
      "4/4 - 0s - loss: 24.9089 - val_loss: 18.4026\n",
      "Epoch 75/100\n",
      "4/4 - 0s - loss: 23.8115 - val_loss: 31.1159\n",
      "Epoch 76/100\n",
      "4/4 - 0s - loss: 37.9183 - val_loss: 18.2472\n",
      "Epoch 77/100\n",
      "4/4 - 0s - loss: 21.6977 - val_loss: 19.4103\n",
      "Epoch 78/100\n",
      "4/4 - 0s - loss: 23.0944 - val_loss: 21.5010\n",
      "Epoch 79/100\n",
      "4/4 - 0s - loss: 20.5002 - val_loss: 17.7216\n",
      "Epoch 80/100\n",
      "4/4 - 0s - loss: 20.0977 - val_loss: 19.6965\n",
      "Epoch 81/100\n",
      "4/4 - 0s - loss: 20.2215 - val_loss: 17.4856\n",
      "Epoch 82/100\n",
      "4/4 - 0s - loss: 20.1904 - val_loss: 26.1668\n",
      "Epoch 83/100\n",
      "4/4 - 0s - loss: 25.6081 - val_loss: 17.8637\n",
      "Epoch 84/100\n",
      "4/4 - 0s - loss: 23.1049 - val_loss: 17.6387\n",
      "Epoch 85/100\n",
      "4/4 - 0s - loss: 21.4030 - val_loss: 19.5475\n",
      "Epoch 86/100\n",
      "4/4 - 0s - loss: 24.3982 - val_loss: 70.6357\n",
      "Epoch 87/100\n",
      "4/4 - 0s - loss: 52.9279 - val_loss: 25.3048\n",
      "Epoch 88/100\n",
      "4/4 - 0s - loss: 25.9008 - val_loss: 17.1101\n",
      "Epoch 89/100\n",
      "4/4 - 0s - loss: 23.5897 - val_loss: 17.5110\n",
      "Epoch 90/100\n",
      "4/4 - 0s - loss: 20.5374 - val_loss: 20.8609\n",
      "Epoch 91/100\n",
      "4/4 - 0s - loss: 24.3780 - val_loss: 43.4734\n",
      "Epoch 92/100\n",
      "4/4 - 0s - loss: 28.7404 - val_loss: 20.0957\n",
      "Epoch 93/100\n",
      "4/4 - 0s - loss: 20.7841 - val_loss: 17.6216\n",
      "Epoch 94/100\n",
      "4/4 - 0s - loss: 19.3086 - val_loss: 20.8584\n",
      "Epoch 95/100\n",
      "4/4 - 0s - loss: 22.6361 - val_loss: 16.8696\n",
      "Epoch 96/100\n",
      "4/4 - 0s - loss: 21.5056 - val_loss: 17.9861\n",
      "Epoch 97/100\n",
      "4/4 - 0s - loss: 19.5062 - val_loss: 19.0513\n",
      "Epoch 98/100\n",
      "4/4 - 0s - loss: 20.0054 - val_loss: 18.6586\n",
      "Epoch 99/100\n",
      "4/4 - 0s - loss: 20.9089 - val_loss: 16.3646\n",
      "Epoch 100/100\n",
      "4/4 - 0s - loss: 19.7602 - val_loss: 21.4056\n",
      "2/2 [==============================] - 0s 2ms/step - loss: 28.1958\n",
      "\n",
      "Test Loss -> 28.20\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "Nin = 13\n",
    "Nh = 5\n",
    "Nout = 1\n",
    "\n",
    "model = ANN(Nin, Nh, Nout)\n",
    "(X_train, y_train), (X_test, y_test) = Data_func()\n",
    "history = model.fit(X_train, y_train, epochs=100, \n",
    "                    batch_size=100, validation_split=0.2, verbose=2)\n",
    "\n",
    "performace_test = model.evaluate(X_test, y_test, batch_size=100)\n",
    "print('\\nTest Loss -> {:.2f}'.format(performace_test))\n",
    "\n",
    "plot_loss(history)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "### 2.3.5 전체코드"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "4/4 - 0s - loss: 446.5541 - val_loss: 128.0201\n",
      "Epoch 2/100\n",
      "4/4 - 0s - loss: 108.9136 - val_loss: 112.4833\n",
      "Epoch 3/100\n",
      "4/4 - 0s - loss: 88.8319 - val_loss: 136.5690\n",
      "Epoch 4/100\n",
      "4/4 - 0s - loss: 76.6614 - val_loss: 68.0886\n",
      "Epoch 5/100\n",
      "4/4 - 0s - loss: 61.7409 - val_loss: 82.5634\n",
      "Epoch 6/100\n",
      "4/4 - 0s - loss: 72.0130 - val_loss: 97.3553\n",
      "Epoch 7/100\n",
      "4/4 - 0s - loss: 55.9550 - val_loss: 54.9127\n",
      "Epoch 8/100\n",
      "4/4 - 0s - loss: 49.9966 - val_loss: 56.6309\n",
      "Epoch 9/100\n",
      "4/4 - 0s - loss: 51.6878 - val_loss: 56.1759\n",
      "Epoch 10/100\n",
      "4/4 - 0s - loss: 65.9424 - val_loss: 46.8505\n",
      "Epoch 11/100\n",
      "4/4 - 0s - loss: 43.2241 - val_loss: 48.1130\n",
      "Epoch 12/100\n",
      "4/4 - 0s - loss: 41.0303 - val_loss: 56.6143\n",
      "Epoch 13/100\n",
      "4/4 - 0s - loss: 68.8754 - val_loss: 40.9030\n",
      "Epoch 14/100\n",
      "4/4 - 0s - loss: 38.7080 - val_loss: 45.9900\n",
      "Epoch 15/100\n",
      "4/4 - 0s - loss: 47.9596 - val_loss: 57.0690\n",
      "Epoch 16/100\n",
      "4/4 - 0s - loss: 43.1500 - val_loss: 48.8059\n",
      "Epoch 17/100\n",
      "4/4 - 0s - loss: 52.3103 - val_loss: 56.6444\n",
      "Epoch 18/100\n",
      "4/4 - 0s - loss: 41.2641 - val_loss: 37.1367\n",
      "Epoch 19/100\n",
      "4/4 - 0s - loss: 35.1449 - val_loss: 40.8880\n",
      "Epoch 20/100\n",
      "4/4 - 0s - loss: 34.7154 - val_loss: 31.1009\n",
      "Epoch 21/100\n",
      "4/4 - 0s - loss: 31.2326 - val_loss: 42.7145\n",
      "Epoch 22/100\n",
      "4/4 - 0s - loss: 53.1995 - val_loss: 46.2895\n",
      "Epoch 23/100\n",
      "4/4 - 0s - loss: 49.0476 - val_loss: 28.6142\n",
      "Epoch 24/100\n",
      "4/4 - 0s - loss: 29.3067 - val_loss: 42.1676\n",
      "Epoch 25/100\n",
      "4/4 - 0s - loss: 29.9319 - val_loss: 30.4918\n",
      "Epoch 26/100\n",
      "4/4 - 0s - loss: 35.1306 - val_loss: 32.2523\n",
      "Epoch 27/100\n",
      "4/4 - 0s - loss: 42.8767 - val_loss: 26.2149\n",
      "Epoch 28/100\n",
      "4/4 - 0s - loss: 26.6557 - val_loss: 33.6422\n",
      "Epoch 29/100\n",
      "4/4 - 0s - loss: 28.1253 - val_loss: 38.9067\n",
      "Epoch 30/100\n",
      "4/4 - 0s - loss: 31.5021 - val_loss: 37.8184\n",
      "Epoch 31/100\n",
      "4/4 - 0s - loss: 30.9841 - val_loss: 47.3701\n",
      "Epoch 32/100\n",
      "4/4 - 0s - loss: 40.6401 - val_loss: 68.7294\n",
      "Epoch 33/100\n",
      "4/4 - 0s - loss: 46.3122 - val_loss: 79.0091\n",
      "Epoch 34/100\n",
      "4/4 - 0s - loss: 56.9219 - val_loss: 27.4362\n",
      "Epoch 35/100\n",
      "4/4 - 0s - loss: 25.9551 - val_loss: 34.3032\n",
      "Epoch 36/100\n",
      "4/4 - 0s - loss: 26.3973 - val_loss: 27.5284\n",
      "Epoch 37/100\n",
      "4/4 - 0s - loss: 24.4472 - val_loss: 26.1327\n",
      "Epoch 38/100\n",
      "4/4 - 0s - loss: 26.0202 - val_loss: 30.2060\n",
      "Epoch 39/100\n",
      "4/4 - 0s - loss: 33.8099 - val_loss: 74.5085\n",
      "Epoch 40/100\n",
      "4/4 - 0s - loss: 56.3971 - val_loss: 28.7124\n",
      "Epoch 41/100\n",
      "4/4 - 0s - loss: 24.3110 - val_loss: 51.6860\n",
      "Epoch 42/100\n",
      "4/4 - 0s - loss: 60.8036 - val_loss: 23.7468\n",
      "Epoch 43/100\n",
      "4/4 - 0s - loss: 25.1669 - val_loss: 47.2875\n",
      "Epoch 44/100\n",
      "4/4 - 0s - loss: 44.7033 - val_loss: 30.2147\n",
      "Epoch 45/100\n",
      "4/4 - 0s - loss: 32.8225 - val_loss: 25.4664\n",
      "Epoch 46/100\n",
      "4/4 - 0s - loss: 22.8016 - val_loss: 22.1441\n",
      "Epoch 47/100\n",
      "4/4 - 0s - loss: 23.3751 - val_loss: 21.7806\n",
      "Epoch 48/100\n",
      "4/4 - 0s - loss: 24.5018 - val_loss: 25.3743\n",
      "Epoch 49/100\n",
      "4/4 - 0s - loss: 28.6978 - val_loss: 22.2976\n",
      "Epoch 50/100\n",
      "4/4 - 0s - loss: 21.5259 - val_loss: 28.6661\n",
      "Epoch 51/100\n",
      "4/4 - 0s - loss: 28.5845 - val_loss: 21.5481\n",
      "Epoch 52/100\n",
      "4/4 - 0s - loss: 23.7620 - val_loss: 43.7886\n",
      "Epoch 53/100\n",
      "4/4 - 0s - loss: 31.1915 - val_loss: 22.6694\n",
      "Epoch 54/100\n",
      "4/4 - 0s - loss: 22.8346 - val_loss: 23.9655\n",
      "Epoch 55/100\n",
      "4/4 - 0s - loss: 21.9163 - val_loss: 21.5680\n",
      "Epoch 56/100\n",
      "4/4 - 0s - loss: 21.5389 - val_loss: 38.7196\n",
      "Epoch 57/100\n",
      "4/4 - 0s - loss: 31.2250 - val_loss: 20.6255\n",
      "Epoch 58/100\n",
      "4/4 - 0s - loss: 21.4629 - val_loss: 21.8689\n",
      "Epoch 59/100\n",
      "4/4 - 0s - loss: 21.2652 - val_loss: 20.5217\n",
      "Epoch 60/100\n",
      "4/4 - 0s - loss: 20.9172 - val_loss: 20.1550\n",
      "Epoch 61/100\n",
      "4/4 - 0s - loss: 20.7546 - val_loss: 20.9083\n",
      "Epoch 62/100\n",
      "4/4 - 0s - loss: 22.5064 - val_loss: 22.1981\n",
      "Epoch 63/100\n",
      "4/4 - 0s - loss: 27.9327 - val_loss: 20.5295\n",
      "Epoch 64/100\n",
      "4/4 - 0s - loss: 20.5493 - val_loss: 24.3155\n",
      "Epoch 65/100\n",
      "4/4 - 0s - loss: 21.0350 - val_loss: 21.3683\n",
      "Epoch 66/100\n",
      "4/4 - 0s - loss: 20.3482 - val_loss: 20.0972\n",
      "Epoch 67/100\n",
      "4/4 - 0s - loss: 20.0919 - val_loss: 24.2947\n",
      "Epoch 68/100\n",
      "4/4 - 0s - loss: 26.4803 - val_loss: 20.6480\n",
      "Epoch 69/100\n",
      "4/4 - 0s - loss: 20.0827 - val_loss: 29.3998\n",
      "Epoch 70/100\n",
      "4/4 - 0s - loss: 29.8785 - val_loss: 32.2786\n",
      "Epoch 71/100\n",
      "4/4 - 0s - loss: 27.9189 - val_loss: 20.5604\n",
      "Epoch 72/100\n",
      "4/4 - 0s - loss: 25.9339 - val_loss: 41.9594\n",
      "Epoch 73/100\n",
      "4/4 - 0s - loss: 35.9592 - val_loss: 23.5539\n",
      "Epoch 74/100\n",
      "4/4 - 0s - loss: 20.7169 - val_loss: 20.2604\n",
      "Epoch 75/100\n",
      "4/4 - 0s - loss: 23.7169 - val_loss: 21.9079\n",
      "Epoch 76/100\n",
      "4/4 - 0s - loss: 22.2018 - val_loss: 19.7816\n",
      "Epoch 77/100\n",
      "4/4 - 0s - loss: 19.8550 - val_loss: 19.1146\n",
      "Epoch 78/100\n",
      "4/4 - 0s - loss: 20.1783 - val_loss: 35.5470\n",
      "Epoch 79/100\n",
      "4/4 - 0s - loss: 34.7747 - val_loss: 23.5279\n",
      "Epoch 80/100\n",
      "4/4 - 0s - loss: 22.0847 - val_loss: 20.5316\n",
      "Epoch 81/100\n",
      "4/4 - 0s - loss: 20.3906 - val_loss: 24.8046\n",
      "Epoch 82/100\n",
      "4/4 - 0s - loss: 27.8631 - val_loss: 18.7018\n",
      "Epoch 83/100\n",
      "4/4 - 0s - loss: 23.1915 - val_loss: 50.7002\n",
      "Epoch 84/100\n",
      "4/4 - 0s - loss: 37.4869 - val_loss: 18.7866\n",
      "Epoch 85/100\n",
      "4/4 - 0s - loss: 19.7619 - val_loss: 41.7948\n",
      "Epoch 86/100\n",
      "4/4 - 0s - loss: 29.7331 - val_loss: 18.9708\n",
      "Epoch 87/100\n",
      "4/4 - 0s - loss: 19.5656 - val_loss: 18.9901\n",
      "Epoch 88/100\n",
      "4/4 - 0s - loss: 20.2034 - val_loss: 22.4540\n",
      "Epoch 89/100\n",
      "4/4 - 0s - loss: 20.1789 - val_loss: 18.6207\n",
      "Epoch 90/100\n",
      "4/4 - 0s - loss: 19.2017 - val_loss: 18.3789\n",
      "Epoch 91/100\n",
      "4/4 - 0s - loss: 23.8929 - val_loss: 18.5436\n",
      "Epoch 92/100\n",
      "4/4 - 0s - loss: 19.2401 - val_loss: 21.7244\n",
      "Epoch 93/100\n",
      "4/4 - 0s - loss: 24.6765 - val_loss: 18.3390\n",
      "Epoch 94/100\n",
      "4/4 - 0s - loss: 19.1304 - val_loss: 18.3038\n",
      "Epoch 95/100\n",
      "4/4 - 0s - loss: 19.0200 - val_loss: 23.2290\n",
      "Epoch 96/100\n",
      "4/4 - 0s - loss: 21.4478 - val_loss: 26.3381\n",
      "Epoch 97/100\n",
      "4/4 - 0s - loss: 21.6271 - val_loss: 18.4226\n",
      "Epoch 98/100\n",
      "4/4 - 0s - loss: 21.2052 - val_loss: 17.6381\n",
      "Epoch 99/100\n",
      "4/4 - 0s - loss: 20.4132 - val_loss: 19.5611\n",
      "Epoch 100/100\n",
      "4/4 - 0s - loss: 19.4064 - val_loss: 18.5854\n",
      "2/2 [==============================] - 0s 2ms/step - loss: 16.3441\n",
      "\n",
      "Test Loss -> 16.34\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# file: ex2_2_ann_rg.py\n",
    "\n",
    "# 1. 회귀 ANN 모델링 \n",
    "from keras import layers, models\n",
    "\n",
    "class ANN(models.Model):\n",
    "    def __init__(self, Nin, Nh, Nout):\n",
    "        # Prepare network layers and activate functions\n",
    "        hidden = layers.Dense(Nh)\n",
    "        output = layers.Dense(Nout)\n",
    "        relu = layers.Activation('relu')\n",
    "\n",
    "        # Connect network elements\n",
    "        x = layers.Input(shape=(Nin,))\n",
    "        h = relu(hidden(x))\n",
    "        y = output(h)\n",
    "\n",
    "        super().__init__(x, y)\n",
    "        self.compile(loss='mse', optimizer='sgd')\n",
    "\n",
    "# 2. 학습과 평가용 데이터 불러오기\n",
    "from keras import datasets\n",
    "from sklearn import preprocessing\n",
    "\n",
    "def Data_func():\n",
    "    (X_train, y_train), (X_test, y_test) = datasets.boston_housing.load_data()\n",
    "    scaler = preprocessing.MinMaxScaler()\n",
    "    X_train = scaler.fit_transform(X_train)\n",
    "    X_test = scaler.transform(X_test)\n",
    "    return (X_train, y_train), (X_test, y_test)\n",
    "\n",
    "\n",
    "# 3. 회귀 ANN 학습 결과 그래프 구현\n",
    "import matplotlib.pyplot as plt\n",
    "from keraspp.skeras import plot_loss\n",
    "\n",
    "# 4. 회귀 ANN 학습 및 성능 분석\n",
    "def main():\n",
    "    Nin = 13\n",
    "    Nh = 5\n",
    "    Nout = 1\n",
    "\n",
    "    model = ANN(Nin, Nh, Nout)\n",
    "    (X_train, y_train), (X_test, y_test) = Data_func()\n",
    "\n",
    "    history = model.fit(X_train, y_train, epochs=100, batch_size=100, validation_split=0.2, verbose=2)\n",
    "\n",
    "    performace_test = model.evaluate(X_test, y_test, batch_size=100)\n",
    "    print('\\nTest Loss -> {:.2f}'.format(performace_test))\n",
    "\n",
    "    plot_loss(history)\n",
    "    plt.show()\n",
    "\n",
    "main()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "keras-gpu",
   "language": "python",
   "name": "keras-gpu"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  },
  "toc": {
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": true,
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": "block",
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}