Facial Expression Training.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import os"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [],
   "source": [
    "path = 'Final/'\n",
    "files = []\n",
    "# r=root, d=directories, f = files\n",
    "for r, d, f in os.walk(path):\n",
    "    for file in f:\n",
    "        if '.csv' in file:\n",
    "            files.append(os.path.join(r, file))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Final/927_0196_01.csv'"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "files[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>gender</th>\n",
       "      <th>gender_no</th>\n",
       "      <th>output</th>\n",
       "      <th>expression_index</th>\n",
       "      <th>video_name</th>\n",
       "      <th>start_time_in_sec</th>\n",
       "      <th>end_time_in_sec</th>\n",
       "      <th>start_exp</th>\n",
       "      <th>end_exp</th>\n",
       "      <th>text_file</th>\n",
       "      <th>kinect_start_frame</th>\n",
       "      <th>kinect_end_frame</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>F</td>\n",
       "      <td>3</td>\n",
       "      <td>An</td>\n",
       "      <td>1</td>\n",
       "      <td>927_0216_01.MP4</td>\n",
       "      <td>5.356700</td>\n",
       "      <td>9.481139</td>\n",
       "      <td>161</td>\n",
       "      <td>284</td>\n",
       "      <td>20170902_092359_00.txt</td>\n",
       "      <td>217</td>\n",
       "      <td>312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>F</td>\n",
       "      <td>3</td>\n",
       "      <td>An</td>\n",
       "      <td>2</td>\n",
       "      <td>927_0216_01.MP4</td>\n",
       "      <td>9.883585</td>\n",
       "      <td>14.903150</td>\n",
       "      <td>296</td>\n",
       "      <td>447</td>\n",
       "      <td>20170902_092359_00.txt</td>\n",
       "      <td>322</td>\n",
       "      <td>466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>F</td>\n",
       "      <td>3</td>\n",
       "      <td>An</td>\n",
       "      <td>3</td>\n",
       "      <td>927_0216_01.MP4</td>\n",
       "      <td>14.903150</td>\n",
       "      <td>19.427206</td>\n",
       "      <td>447</td>\n",
       "      <td>582</td>\n",
       "      <td>20170902_092359_00.txt</td>\n",
       "      <td>466</td>\n",
       "      <td>564</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>F</td>\n",
       "      <td>3</td>\n",
       "      <td>An</td>\n",
       "      <td>4</td>\n",
       "      <td>927_0216_01.MP4</td>\n",
       "      <td>19.427206</td>\n",
       "      <td>25.077029</td>\n",
       "      <td>582</td>\n",
       "      <td>752</td>\n",
       "      <td>20170902_092359_00.txt</td>\n",
       "      <td>564</td>\n",
       "      <td>699</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>F</td>\n",
       "      <td>3</td>\n",
       "      <td>An</td>\n",
       "      <td>5</td>\n",
       "      <td>927_0216_01.MP4</td>\n",
       "      <td>25.077029</td>\n",
       "      <td>29.981168</td>\n",
       "      <td>752</td>\n",
       "      <td>899</td>\n",
       "      <td>20170902_092359_00.txt</td>\n",
       "      <td>699</td>\n",
       "      <td>828</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  gender  gender_no output  expression_index       video_name  \\\n",
       "1      F          3     An                 1  927_0216_01.MP4   \n",
       "2      F          3     An                 2  927_0216_01.MP4   \n",
       "3      F          3     An                 3  927_0216_01.MP4   \n",
       "4      F          3     An                 4  927_0216_01.MP4   \n",
       "5      F          3     An                 5  927_0216_01.MP4   \n",
       "\n",
       "   start_time_in_sec  end_time_in_sec  start_exp  end_exp  \\\n",
       "1           5.356700         9.481139        161      284   \n",
       "2           9.883585        14.903150        296      447   \n",
       "3          14.903150        19.427206        447      582   \n",
       "4          19.427206        25.077029        582      752   \n",
       "5          25.077029        29.981168        752      899   \n",
       "\n",
       "                text_file  kinect_start_frame  kinect_end_frame  \n",
       "1  20170902_092359_00.txt                 217               312  \n",
       "2  20170902_092359_00.txt                 322               466  \n",
       "3  20170902_092359_00.txt                 466               564  \n",
       "4  20170902_092359_00.txt                 564               699  \n",
       "5  20170902_092359_00.txt                 699               828  "
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "descriptors = pd.read_csv('Descriptors.csv', index_col=0)\n",
    "descriptors.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_file(file):\n",
    "    temp_df = pd.read_csv(file, sep=',')\n",
    "    max_value = np.amax(temp_df)\n",
    "    max_value = max(max_value)\n",
    "    temp_df = temp_df/max_value\n",
    "    #print(max_value)\n",
    "    name = file[6:-4]\n",
    "    output_value = descriptors.loc[descriptors['video_name']==name +'.MP4']['output'][:1].values[0]\n",
    "    #video_name = descriptors.loc[descriptors['text_file']==textfile]['video_name'][:1].values[0]\n",
    "    gender = descriptors.loc[descriptors['video_name']==name +'.MP4']['gender'][:1].values[0]\n",
    "    #temp_df['textfile'] = textfile\n",
    "    #temp_df['output'] = output_value\n",
    "    #temp_df['video_name'] = video_name\n",
    "    if gender == 'F':\n",
    "        temp_df['gender'] = 1\n",
    "    else:\n",
    "        temp_df['gender'] = 0\n",
    "    #temp_df.drop(['Unnamed: 152'], axis=1, inplace=True)\n",
    "    return temp_df.values, output_value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [],
   "source": [
    "a, b = load_file('Final/927_0196_01.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(791, 2279)"
      ]
     },
     "execution_count": 129,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [],
   "source": [
    "shapes = []\n",
    "Y = []\n",
    "for name in files:\n",
    "    data, t = load_file(name)\n",
    "    shapes.append(data.shape[0])\n",
    "    Y.append(t)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [],
   "source": [
    "index = 0 \n",
    "for element in Y:\n",
    "    if element == 'Ne':\n",
    "        Y[index] = 0\n",
    "    elif element == 'Sa':\n",
    "        Y[index] = 1\n",
    "    elif element == 'Su':\n",
    "        Y[index] = 2\n",
    "    elif element == 'Fe':\n",
    "        Y[index] = 3\n",
    "    elif element == 'An':\n",
    "        Y[index] = 4\n",
    "    elif element == 'Di':\n",
    "        Y[index] = 5\n",
    "    elif element == 'Ha':\n",
    "        Y[index] = 6\n",
    "    index += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1864\n"
     ]
    }
   ],
   "source": [
    "from tensorflow.keras.utils import to_categorical\n",
    "\n",
    "maxlen = max(shapes)\n",
    "print(maxlen)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_group(filenames, prefix=''):\n",
    "    loaded = list()\n",
    "    for name in filenames:\n",
    "        data, temp = load_file(name)\n",
    "        #print('The name is:', name, data.shape)\n",
    "        zeros_data = np.zeros((max(shapes)-data.shape[0], 2279))\n",
    "        print(data.shape)\n",
    "        data = np.concatenate((zeros_data, data), axis=0)\n",
    "        #print('Now the shape is:', name, data.shape)\n",
    "        loaded.append(data)\n",
    "        shapes.append(data.shape[0])\n",
    "    loaded = np.stack(loaded)\n",
    "    return loaded"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(791, 2279)\n",
      "(1067, 2279)\n",
      "(1112, 2279)\n",
      "(1362, 2279)\n",
      "(965, 2279)\n",
      "(1252, 2279)\n",
      "(1027, 2279)\n",
      "(1035, 2279)\n",
      "(1419, 2279)\n",
      "(1321, 2279)\n",
      "(1351, 2279)\n",
      "(952, 2279)\n",
      "(1310, 2279)\n",
      "(907, 2279)\n",
      "(914, 2279)\n",
      "(1131, 2279)\n",
      "(972, 2279)\n",
      "(919, 2279)\n",
      "(1083, 2279)\n",
      "(1046, 2279)\n",
      "(1022, 2279)\n",
      "(842, 2279)\n",
      "(950, 2279)\n",
      "(1265, 2279)\n",
      "(796, 2279)\n",
      "(824, 2279)\n",
      "(781, 2279)\n",
      "(738, 2279)\n",
      "(895, 2279)\n",
      "(1124, 2279)\n",
      "(1021, 2279)\n",
      "(1174, 2279)\n",
      "(852, 2279)\n",
      "(869, 2279)\n",
      "(723, 2279)\n",
      "(792, 2279)\n",
      "(758, 2279)\n",
      "(775, 2279)\n",
      "(742, 2279)\n",
      "(742, 2279)\n",
      "(1605, 2279)\n",
      "(575, 2279)\n",
      "(836, 2279)\n",
      "(1147, 2279)\n",
      "(1105, 2279)\n",
      "(1174, 2279)\n",
      "(733, 2279)\n",
      "(1165, 2279)\n",
      "(682, 2279)\n",
      "(1060, 2279)\n",
      "(1097, 2279)\n",
      "(949, 2279)\n",
      "(1295, 2279)\n",
      "(841, 2279)\n",
      "(1182, 2279)\n",
      "(958, 2279)\n",
      "(1864, 2279)\n",
      "(1577, 2279)\n",
      "(1150, 2279)\n",
      "(1224, 2279)\n",
      "(887, 2279)\n",
      "(1332, 2279)\n",
      "(1055, 2279)\n",
      "(63, 1864, 2279) (63, 7)\n"
     ]
    }
   ],
   "source": [
    "X = load_group(files)\n",
    "\n",
    "Y = np.asarray(Y).reshape((63,1))\n",
    "Y = to_categorical(Y)\n",
    "print(X.shape, Y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.save('X.npy', X)\n",
    "np.save('Y.npy', Y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_test, Y_train, Y_test = train_test_split(X,Y, test_size = 0.1, random_state = 42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense, Dropout, LSTM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [],
   "source": [
    "def evaluate_model(trainX, trainy, testX, testy):\n",
    "    verbose = 1\n",
    "    epochs = 100\n",
    "    n_timesteps, n_features, n_outputs = trainX.shape[1], trainX.shape[2], trainy.shape[1]\n",
    "    model = Sequential()\n",
    "    model.add(LSTM(32, input_shape=(n_timesteps, n_features)))\n",
    "    #model.add(Dropout(0.5))\n",
    "    model.add(Dense(16, activation='relu'))\n",
    "    model.add(Dense(n_outputs, activation='softmax'))\n",
    "    model.compile(loss='categorical_crossentropy', optimizer = 'adam', metrics = ['accuracy'])\n",
    "    model.fit(trainX, trainy, epochs=epochs, verbose=verbose)\n",
    "    _, accuracy = model.evaluate(testX, testy, verbose=verbose)\n",
    "    return accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "56/56 [==============================] - 36s 637ms/sample - loss: 2.0371 - accuracy: 0.1071\n",
      "Epoch 2/100\n",
      "56/56 [==============================] - 16s 278ms/sample - loss: 1.9467 - accuracy: 0.1429\n",
      "Epoch 3/100\n",
      "56/56 [==============================] - 12s 212ms/sample - loss: 1.9439 - accuracy: 0.1607\n",
      "Epoch 4/100\n",
      "56/56 [==============================] - 10s 183ms/sample - loss: 1.9432 - accuracy: 0.1607\n",
      "Epoch 5/100\n",
      "56/56 [==============================] - 10s 179ms/sample - loss: 1.9437 - accuracy: 0.1607\n",
      "Epoch 6/100\n",
      "56/56 [==============================] - 10s 175ms/sample - loss: 1.9420 - accuracy: 0.1607\n",
      "Epoch 7/100\n",
      "56/56 [==============================] - 10s 176ms/sample - loss: 1.9416 - accuracy: 0.1607\n",
      "Epoch 8/100\n",
      "56/56 [==============================] - 10s 177ms/sample - loss: 1.9420 - accuracy: 0.1607\n",
      "Epoch 9/100\n",
      "56/56 [==============================] - 10s 175ms/sample - loss: 1.9413 - accuracy: 0.1607\n",
      "Epoch 10/100\n",
      "56/56 [==============================] - 10s 176ms/sample - loss: 1.9422 - accuracy: 0.1607\n",
      "Epoch 11/100\n",
      "56/56 [==============================] - 10s 182ms/sample - loss: 1.9415 - accuracy: 0.1607\n",
      "Epoch 12/100\n",
      "56/56 [==============================] - 10s 184ms/sample - loss: 1.9410 - accuracy: 0.1607\n",
      "Epoch 13/100\n",
      "56/56 [==============================] - 10s 183ms/sample - loss: 1.9413 - accuracy: 0.1607\n",
      "Epoch 14/100\n",
      "56/56 [==============================] - 11s 195ms/sample - loss: 1.9412 - accuracy: 0.1607\n",
      "Epoch 15/100\n",
      "56/56 [==============================] - 12s 218ms/sample - loss: 1.9410 - accuracy: 0.1607\n",
      "Epoch 16/100\n",
      "56/56 [==============================] - 13s 227ms/sample - loss: 1.9412 - accuracy: 0.1607\n",
      "Epoch 17/100\n",
      "56/56 [==============================] - 13s 234ms/sample - loss: 1.9424 - accuracy: 0.1607\n",
      "Epoch 18/100\n",
      "56/56 [==============================] - 12s 209ms/sample - loss: 1.9403 - accuracy: 0.1607\n",
      "Epoch 19/100\n",
      "56/56 [==============================] - 13s 228ms/sample - loss: 1.9406 - accuracy: 0.1607\n",
      "Epoch 20/100\n",
      "56/56 [==============================] - 13s 229ms/sample - loss: 1.9403 - accuracy: 0.1607\n",
      "Epoch 21/100\n",
      "56/56 [==============================] - 13s 230ms/sample - loss: 1.9404 - accuracy: 0.1607\n",
      "Epoch 22/100\n",
      "56/56 [==============================] - 13s 230ms/sample - loss: 1.9411 - accuracy: 0.1607\n",
      "Epoch 23/100\n",
      "56/56 [==============================] - 12s 223ms/sample - loss: 1.9400 - accuracy: 0.1607\n",
      "Epoch 24/100\n",
      "56/56 [==============================] - 12s 223ms/sample - loss: 1.9400 - accuracy: 0.1607\n",
      "Epoch 25/100\n",
      "56/56 [==============================] - 12s 221ms/sample - loss: 1.9402 - accuracy: 0.1607\n",
      "Epoch 26/100\n",
      "56/56 [==============================] - 12s 220ms/sample - loss: 1.9403 - accuracy: 0.1607\n",
      "Epoch 27/100\n",
      "56/56 [==============================] - 13s 233ms/sample - loss: 1.9395 - accuracy: 0.1607\n",
      "Epoch 28/100\n",
      "56/56 [==============================] - 14s 245ms/sample - loss: 1.9392 - accuracy: 0.1607\n",
      "Epoch 29/100\n",
      "56/56 [==============================] - 13s 235ms/sample - loss: 1.9393 - accuracy: 0.1607\n",
      "Epoch 30/100\n",
      "56/56 [==============================] - 13s 224ms/sample - loss: 1.9388 - accuracy: 0.1607\n",
      "Epoch 31/100\n",
      "56/56 [==============================] - 12s 216ms/sample - loss: 1.9379 - accuracy: 0.1607\n",
      "Epoch 32/100\n",
      "56/56 [==============================] - 212s 4s/sample - loss: 1.9382 - accuracy: 0.1607\n",
      "Epoch 33/100\n",
      "56/56 [==============================] - 68s 1s/sample - loss: 1.9374 - accuracy: 0.1607\n",
      "Epoch 34/100\n",
      "56/56 [==============================] - 40s 708ms/sample - loss: 1.9381 - accuracy: 0.1607\n",
      "Epoch 35/100\n",
      "56/56 [==============================] - 20s 351ms/sample - loss: 1.9370 - accuracy: 0.1607\n",
      "Epoch 36/100\n",
      "56/56 [==============================] - 20s 354ms/sample - loss: 1.9360 - accuracy: 0.1607\n",
      "Epoch 37/100\n",
      "56/56 [==============================] - 20s 350ms/sample - loss: 1.9353 - accuracy: 0.1607\n",
      "Epoch 38/100\n",
      "56/56 [==============================] - 20s 358ms/sample - loss: 1.9348 - accuracy: 0.1607\n",
      "Epoch 39/100\n",
      "56/56 [==============================] - 20s 357ms/sample - loss: 1.9341 - accuracy: 0.1607\n",
      "Epoch 40/100\n",
      "56/56 [==============================] - 21s 369ms/sample - loss: 1.9339 - accuracy: 0.1607\n",
      "Epoch 41/100\n",
      "56/56 [==============================] - 20s 366ms/sample - loss: 1.9335 - accuracy: 0.1607\n",
      "Epoch 42/100\n",
      "56/56 [==============================] - 21s 369ms/sample - loss: 1.9334 - accuracy: 0.1607\n",
      "Epoch 43/100\n",
      "56/56 [==============================] - 20s 357ms/sample - loss: 1.9338 - accuracy: 0.1607\n",
      "Epoch 44/100\n",
      "56/56 [==============================] - 19s 333ms/sample - loss: 1.9313 - accuracy: 0.1607\n",
      "Epoch 45/100\n",
      "56/56 [==============================] - 19s 342ms/sample - loss: 1.9344 - accuracy: 0.1607\n",
      "Epoch 46/100\n",
      "56/56 [==============================] - 18s 319ms/sample - loss: 1.9296 - accuracy: 0.1607\n",
      "Epoch 47/100\n",
      "56/56 [==============================] - 18s 329ms/sample - loss: 1.9300 - accuracy: 0.1607\n",
      "Epoch 48/100\n",
      "56/56 [==============================] - 17s 308ms/sample - loss: 1.9267 - accuracy: 0.1607\n",
      "Epoch 49/100\n",
      "56/56 [==============================] - 17s 300ms/sample - loss: 1.9286 - accuracy: 0.1607\n",
      "Epoch 50/100\n",
      "56/56 [==============================] - 20s 351ms/sample - loss: 1.9250 - accuracy: 0.1964\n",
      "Epoch 51/100\n",
      "56/56 [==============================] - 19s 348ms/sample - loss: 1.9198 - accuracy: 0.2143\n",
      "Epoch 52/100\n",
      "56/56 [==============================] - 19s 342ms/sample - loss: 1.9247 - accuracy: 0.2143\n",
      "Epoch 53/100\n",
      "56/56 [==============================] - 20s 351ms/sample - loss: 1.9126 - accuracy: 0.3214\n",
      "Epoch 54/100\n",
      "56/56 [==============================] - 20s 358ms/sample - loss: 1.9226 - accuracy: 0.1964\n",
      "Epoch 55/100\n",
      "56/56 [==============================] - 20s 357ms/sample - loss: 1.9232 - accuracy: 0.1607\n",
      "Epoch 56/100\n",
      "56/56 [==============================] - 20s 358ms/sample - loss: 1.9187 - accuracy: 0.1786\n",
      "Epoch 57/100\n",
      "56/56 [==============================] - 20s 349ms/sample - loss: 1.9195 - accuracy: 0.2500\n",
      "Epoch 58/100\n",
      "56/56 [==============================] - 20s 363ms/sample - loss: 1.9200 - accuracy: 0.2857\n",
      "Epoch 59/100\n",
      "56/56 [==============================] - 20s 360ms/sample - loss: 1.9179 - accuracy: 0.2321\n",
      "Epoch 60/100\n",
      "56/56 [==============================] - 20s 357ms/sample - loss: 1.9135 - accuracy: 0.2500\n",
      "Epoch 61/100\n",
      "56/56 [==============================] - 20s 355ms/sample - loss: 1.9106 - accuracy: 0.2857\n",
      "Epoch 62/100\n",
      "56/56 [==============================] - 21s 367ms/sample - loss: 1.9052 - accuracy: 0.3214\n",
      "Epoch 63/100\n",
      "56/56 [==============================] - 21s 371ms/sample - loss: 1.9095 - accuracy: 0.2679\n",
      "Epoch 64/100\n",
      "56/56 [==============================] - 21s 373ms/sample - loss: 1.9085 - accuracy: 0.1786\n",
      "Epoch 65/100\n",
      "56/56 [==============================] - 20s 364ms/sample - loss: 1.9063 - accuracy: 0.3036\n",
      "Epoch 66/100\n",
      "56/56 [==============================] - 20s 361ms/sample - loss: 1.9045 - accuracy: 0.2500\n",
      "Epoch 67/100\n",
      "56/56 [==============================] - 20s 351ms/sample - loss: 1.9000 - accuracy: 0.2857\n",
      "Epoch 68/100\n",
      "56/56 [==============================] - 20s 362ms/sample - loss: 1.9004 - accuracy: 0.2679\n",
      "Epoch 69/100\n",
      "56/56 [==============================] - 20s 354ms/sample - loss: 1.8929 - accuracy: 0.3036\n",
      "Epoch 70/100\n",
      "56/56 [==============================] - 20s 363ms/sample - loss: 1.8902 - accuracy: 0.3393\n",
      "Epoch 71/100\n",
      "56/56 [==============================] - 20s 359ms/sample - loss: 1.8851 - accuracy: 0.3571\n",
      "Epoch 72/100\n",
      "56/56 [==============================] - 20s 360ms/sample - loss: 1.8840 - accuracy: 0.3214\n",
      "Epoch 73/100\n",
      "56/56 [==============================] - 20s 364ms/sample - loss: 1.8820 - accuracy: 0.2857\n",
      "Epoch 74/100\n",
      "56/56 [==============================] - 20s 363ms/sample - loss: 1.8792 - accuracy: 0.2857\n",
      "Epoch 75/100\n",
      "56/56 [==============================] - 20s 356ms/sample - loss: 1.8845 - accuracy: 0.2679\n",
      "Epoch 76/100\n",
      "56/56 [==============================] - 20s 359ms/sample - loss: 1.8758 - accuracy: 0.2857\n",
      "Epoch 77/100\n",
      "56/56 [==============================] - 20s 361ms/sample - loss: 1.8742 - accuracy: 0.2857\n",
      "Epoch 78/100\n",
      "56/56 [==============================] - 20s 363ms/sample - loss: 1.8679 - accuracy: 0.3571\n",
      "Epoch 79/100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "56/56 [==============================] - 21s 366ms/sample - loss: 1.8688 - accuracy: 0.3393\n",
      "Epoch 80/100\n",
      "56/56 [==============================] - 20s 360ms/sample - loss: 1.8669 - accuracy: 0.3214\n",
      "Epoch 81/100\n",
      "56/56 [==============================] - 20s 361ms/sample - loss: 1.8627 - accuracy: 0.3214\n",
      "Epoch 82/100\n",
      "56/56 [==============================] - 20s 360ms/sample - loss: 1.8558 - accuracy: 0.3750\n",
      "Epoch 83/100\n",
      "56/56 [==============================] - 21s 368ms/sample - loss: 1.8526 - accuracy: 0.3036\n",
      "Epoch 84/100\n",
      "56/56 [==============================] - 20s 362ms/sample - loss: 1.8584 - accuracy: 0.3214\n",
      "Epoch 85/100\n",
      "56/56 [==============================] - 20s 359ms/sample - loss: 1.8452 - accuracy: 0.3214\n",
      "Epoch 86/100\n",
      "56/56 [==============================] - 20s 361ms/sample - loss: 1.8507 - accuracy: 0.2679\n",
      "Epoch 87/100\n",
      "56/56 [==============================] - 20s 361ms/sample - loss: 1.8468 - accuracy: 0.3393\n",
      "Epoch 88/100\n",
      "56/56 [==============================] - 20s 366ms/sample - loss: 1.8437 - accuracy: 0.3393\n",
      "Epoch 89/100\n",
      "56/56 [==============================] - 20s 358ms/sample - loss: 1.8456 - accuracy: 0.2679\n",
      "Epoch 90/100\n",
      "56/56 [==============================] - 20s 362ms/sample - loss: 1.8372 - accuracy: 0.3036\n",
      "Epoch 91/100\n",
      "56/56 [==============================] - 20s 362ms/sample - loss: 1.8462 - accuracy: 0.3393\n",
      "Epoch 92/100\n",
      "56/56 [==============================] - 20s 358ms/sample - loss: 1.8303 - accuracy: 0.3571\n",
      "Epoch 93/100\n",
      "56/56 [==============================] - 20s 361ms/sample - loss: 1.8297 - accuracy: 0.3036\n",
      "Epoch 94/100\n",
      "56/56 [==============================] - 20s 361ms/sample - loss: 1.8206 - accuracy: 0.3214\n",
      "Epoch 95/100\n",
      "56/56 [==============================] - 20s 358ms/sample - loss: 1.8232 - accuracy: 0.3214\n",
      "Epoch 96/100\n",
      "56/56 [==============================] - 20s 359ms/sample - loss: 1.8114 - accuracy: 0.3036\n",
      "Epoch 97/100\n",
      "56/56 [==============================] - 19s 347ms/sample - loss: 1.8166 - accuracy: 0.3036\n",
      "Epoch 98/100\n",
      "56/56 [==============================] - 20s 361ms/sample - loss: 1.8135 - accuracy: 0.3750\n",
      "Epoch 99/100\n",
      "56/56 [==============================] - 20s 362ms/sample - loss: 1.8088 - accuracy: 0.3750\n",
      "Epoch 100/100\n",
      "56/56 [==============================] - 20s 359ms/sample - loss: 1.8285 - accuracy: 0.3036\n",
      "7/7 [==============================] - 8s 1s/sample - loss: 2.1786 - accuracy: 0.0000e+00\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.0"
      ]
     },
     "execution_count": 138,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "evaluate_model(X_train, Y_train, X_test, Y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.7.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}