stroke-analytics-hackaton.ipynb

{
 "cells": [
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "import pandas as pd\nimport numpy as np\nimport matplotlib.pyplot as plt\n\nimport matplotlib.pyplot as plt\nimport matplotlib.mlab as mlab\nimport seaborn\nfrom pprint import pprint\n\nimport sys\nprint(sys.version)\n\nfrom keras.models import Sequential\nfrom keras.layers import Dense, Dropout, Embedding\nfrom keras.optimizers import Adam\nimport keras.backend as K\nfrom sklearn import preprocessing\nfrom sklearn.metrics import confusion_matrix\nfrom sklearn.ensemble import RandomForestClassifier\nfrom sklearn.ensemble import AdaBoostClassifier\nfrom sklearn.preprocessing import LabelBinarizer\nfrom sklearn import svm\nfrom itertools import product\nimport mxnet as mx\n\n%matplotlib inline",
   "execution_count": 1,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": "3.6.4 |Anaconda custom (64-bit)| (default, Jan 16 2018, 12:04:33) \n[GCC 4.2.1 Compatible Clang 4.0.1 (tags/RELEASE_401/final)]\n"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": "Using TensorFlow backend.\n"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "",
   "execution_count": null,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "TEST_CSV = 'test_v2akXPA.csv'\nTRAIN_CSV = 'train_ajEneEa.csv'\n\n\n\n'''\nbooleans: hypertension, heart_disease, ever_married, stroke\ncategories: gender, work_type, Residence_type, smoking_status\nscalars: age, avg_glucose_level, bmi\n'''\n\ndropColumns = ['gender', 'work_type', 'Residence_type', 'smoking_status', 'stroke']\n\ndef getCleanData(csv):\n    df = pd.read_csv(csv, index_col='id')\n\n    intToBool = {0: False, 1: True}\n    strToBool = {'No': False, 'Yes': True}\n    strToInt = {'No': 0, 'Yes': 1}\n\n    df['gender'] = df['gender'].astype('category')\n    df['work_type'] = df['work_type'].astype('category')\n    df['Residence_type'] = df['Residence_type'].astype('category')\n    df['smoking_status'] = df['smoking_status'].fillna('unknown').astype('category')\n    \n    meanBMI = df['bmi'].mean()\n    df['bmi'] = df['bmi'].fillna(meanBMI)\n    \n#     df['hypertension'] = df['hypertension'].map(intToBool)\n#     df['heart_disease']= df['heart_disease'].map(intToBool)\n#     df['ever_married'] = df['ever_married'].map(strToBool)\n    df['ever_married'] = df['ever_married'].map(strToInt)\n    \n    if 'stroke' in df.columns:\n        df['stroke']= df['stroke'].map(intToBool)\n        \n    categoricalColumns = ['gender', 'work_type', 'Residence_type', 'smoking_status']\n    for column in categoricalColumns:\n        df = df.join(pd.get_dummies(df[column], prefix=column))\n    \n    return df\n\n\n\n\ntrainingData = getCleanData(TRAIN_CSV)\ntestData = getCleanData(TEST_CSV)",
   "execution_count": 2,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "trainingData[:10]",
   "execution_count": 3,
   "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>age</th>\n      <th>hypertension</th>\n      <th>heart_disease</th>\n      <th>ever_married</th>\n      <th>work_type</th>\n      <th>Residence_type</th>\n      <th>avg_glucose_level</th>\n      <th>bmi</th>\n      <th>smoking_status</th>\n      <th>...</th>\n      <th>work_type_Never_worked</th>\n      <th>work_type_Private</th>\n      <th>work_type_Self-employed</th>\n      <th>work_type_children</th>\n      <th>Residence_type_Rural</th>\n      <th>Residence_type_Urban</th>\n      <th>smoking_status_formerly smoked</th>\n      <th>smoking_status_never smoked</th>\n      <th>smoking_status_smokes</th>\n      <th>smoking_status_unknown</th>\n    </tr>\n    <tr>\n      <th>id</th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>30669</th>\n      <td>Male</td>\n      <td>3.0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>children</td>\n      <td>Rural</td>\n      <td>95.12</td>\n      <td>18.0</td>\n      <td>unknown</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>30468</th>\n      <td>Male</td>\n      <td>58.0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>1</td>\n      <td>Private</td>\n      <td>Urban</td>\n      <td>87.96</td>\n      <td>39.2</td>\n      <td>never smoked</td>\n      <td>...</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>16523</th>\n      <td>Female</td>\n      <td>8.0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>Private</td>\n      <td>Urban</td>\n      <td>110.89</td>\n      <td>17.6</td>\n      <td>unknown</td>\n      <td>...</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>56543</th>\n      <td>Female</td>\n      <td>70.0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>Private</td>\n      <td>Rural</td>\n      <td>69.04</td>\n      <td>35.9</td>\n      <td>formerly smoked</td>\n      <td>...</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>46136</th>\n      <td>Male</td>\n      <td>14.0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>Never_worked</td>\n      <td>Rural</td>\n      <td>161.28</td>\n      <td>19.1</td>\n      <td>unknown</td>\n      <td>...</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>32257</th>\n      <td>Female</td>\n      <td>47.0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>Private</td>\n      <td>Urban</td>\n      <td>210.95</td>\n      <td>50.1</td>\n      <td>unknown</td>\n      <td>...</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>52800</th>\n      <td>Female</td>\n      <td>52.0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>Private</td>\n      <td>Urban</td>\n      <td>77.59</td>\n      <td>17.7</td>\n      <td>formerly smoked</td>\n      <td>...</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>41413</th>\n      <td>Female</td>\n      <td>75.0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>1</td>\n      <td>Self-employed</td>\n      <td>Rural</td>\n      <td>243.53</td>\n      <td>27.0</td>\n      <td>never smoked</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>15266</th>\n      <td>Female</td>\n      <td>32.0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>Private</td>\n      <td>Rural</td>\n      <td>77.67</td>\n      <td>32.3</td>\n      <td>smokes</td>\n      <td>...</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>28674</th>\n      <td>Female</td>\n      <td>74.0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>1</td>\n      <td>Self-employed</td>\n      <td>Urban</td>\n      <td>205.84</td>\n      <td>54.6</td>\n      <td>never smoked</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n  </tbody>\n</table>\n<p>10 rows × 25 columns</p>\n</div>",
      "text/plain": "       gender   age  hypertension  heart_disease  ever_married      work_type  \\\nid                                                                              \n30669    Male   3.0             0              0             0       children   \n30468    Male  58.0             1              0             1        Private   \n16523  Female   8.0             0              0             0        Private   \n56543  Female  70.0             0              0             1        Private   \n46136    Male  14.0             0              0             0   Never_worked   \n32257  Female  47.0             0              0             1        Private   \n52800  Female  52.0             0              0             1        Private   \n41413  Female  75.0             0              1             1  Self-employed   \n15266  Female  32.0             0              0             1        Private   \n28674  Female  74.0             1              0             1  Self-employed   \n\n      Residence_type  avg_glucose_level   bmi   smoking_status  \\\nid                                                               \n30669          Rural              95.12  18.0          unknown   \n30468          Urban              87.96  39.2     never smoked   \n16523          Urban             110.89  17.6          unknown   \n56543          Rural              69.04  35.9  formerly smoked   \n46136          Rural             161.28  19.1          unknown   \n32257          Urban             210.95  50.1          unknown   \n52800          Urban              77.59  17.7  formerly smoked   \n41413          Rural             243.53  27.0     never smoked   \n15266          Rural              77.67  32.3           smokes   \n28674          Urban             205.84  54.6     never smoked   \n\n                ...            work_type_Never_worked  work_type_Private  \\\nid              ...                                                        \n30669           ...                                 0                  0   \n30468           ...                                 0                  1   \n16523           ...                                 0                  1   \n56543           ...                                 0                  1   \n46136           ...                                 1                  0   \n32257           ...                                 0                  1   \n52800           ...                                 0                  1   \n41413           ...                                 0                  0   \n15266           ...                                 0                  1   \n28674           ...                                 0                  0   \n\n       work_type_Self-employed  work_type_children  Residence_type_Rural  \\\nid                                                                         \n30669                        0                   1                     1   \n30468                        0                   0                     0   \n16523                        0                   0                     0   \n56543                        0                   0                     1   \n46136                        0                   0                     1   \n32257                        0                   0                     0   \n52800                        0                   0                     0   \n41413                        1                   0                     1   \n15266                        0                   0                     1   \n28674                        1                   0                     0   \n\n       Residence_type_Urban  smoking_status_formerly smoked  \\\nid                                                            \n30669                     0                               0   \n30468                     1                               0   \n16523                     1                               0   \n56543                     0                               1   \n46136                     0                               0   \n32257                     1                               0   \n52800                     1                               1   \n41413                     0                               0   \n15266                     0                               0   \n28674                     1                               0   \n\n       smoking_status_never smoked  smoking_status_smokes  \\\nid                                                          \n30669                            0                      0   \n30468                            1                      0   \n16523                            0                      0   \n56543                            0                      0   \n46136                            0                      0   \n32257                            0                      0   \n52800                            0                      0   \n41413                            1                      0   \n15266                            0                      1   \n28674                            1                      0   \n\n       smoking_status_unknown  \nid                             \n30669                       1  \n30468                       0  \n16523                       1  \n56543                       0  \n46136                       1  \n32257                       1  \n52800                       0  \n41413                       0  \n15266                       0  \n28674                       0  \n\n[10 rows x 25 columns]"
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "trainingData.isnull().any()",
   "execution_count": 4,
   "outputs": [
    {
     "data": {
      "text/plain": "gender                            False\nage                               False\nhypertension                      False\nheart_disease                     False\never_married                      False\nwork_type                         False\nResidence_type                    False\navg_glucose_level                 False\nbmi                               False\nsmoking_status                    False\nstroke                            False\ngender_Female                     False\ngender_Male                       False\ngender_Other                      False\nwork_type_Govt_job                False\nwork_type_Never_worked            False\nwork_type_Private                 False\nwork_type_Self-employed           False\nwork_type_children                False\nResidence_type_Rural              False\nResidence_type_Urban              False\nsmoking_status_formerly smoked    False\nsmoking_status_never smoked       False\nsmoking_status_smokes             False\nsmoking_status_unknown            False\ndtype: bool"
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "df_corr = trainingData.corr()\n\nplt.figure(figsize=(15,10))\nseaborn.heatmap(df_corr)\nseaborn.set(font_scale=2)\nplt.title('Heatmap correlation')\nplt.show()\n",
   "execution_count": 5,
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": "<matplotlib.figure.Figure at 0x1a22d7a860>"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "rank = df_corr['stroke']\ndf_rank = pd.DataFrame(rank)\ndf_rank = np.abs(df_rank).sort_values(by='stroke',ascending=False)\ndf_rank.dropna(inplace=True)\n\ndf_stroke = trainingData[trainingData['stroke'] == True]\nplt.figure(figsize=(15,10))\nplt.scatter(df_stroke['age'], df_stroke['bmi'])\nplt.title('age and bmi corelations to stroke')\nplt.show()\n",
   "execution_count": 6,
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": "<matplotlib.figure.Figure at 0x109c90470>"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "def model():\n    global network_history\n    model = Sequential()\n    model.add(Dense(64,input_shape=(20,)))\n    model.add(Dropout(0.2))\n    model.add(Dense(32, kernel_initializer='normal'))\n    model.add(Dense(1, activation='relu'))\n    \n    model.compile(loss='mean_squared_error', optimizer=Adam(lr=0.0001), metrics=['accuracy'])\n    \n    network_history = model.fit(X_train, y_train, batch_size=128, epochs=600, verbose=1)\n    return model",
   "execution_count": 7,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "\nsplit = int((len(trainingData) - 1) // 0.8)\n\ndf_train_all = df_test_all = trainingData.copy()\n\ndf_train_all = trainingData[0:split]\ndf_test_all = trainingData.sample(split - len(trainingData))\n\nprint(\"training with:\", len(df_train_all), \"rows\")\nprint(\"testing with:\", len(df_test_all))\n\ndf_train_1 = df_train_all[df_train_all['stroke'] == True]\ndf_train_0 = df_train_all[df_train_all['stroke'] == False]\n\nprint(\"training strokes: {} no strokes: {}\".format(len(df_train_1), len(df_train_0)))\n\ndf_sample = df_train_0.sample(783 * 3)\ndf_train = df_train_1.append(df_sample)\ndf_train = df_train.sample(frac=1)\n\ndf_test_1 = df_test_all[df_test_all['stroke'] == True]\ndf_test_0 = df_test_all[df_test_all['stroke'] == False]\n\nprint(\"test strokes: {} no strokes: {}\".format(len(df_test_1), len(df_test_0)))\n\ndf_sample_test = df_test_0.sample(184 * 3)\ndf_test = df_test_1.append(df_sample_test)\ndf_test = df_test.sample(frac=1)\n\nX_train = df_train.drop(dropColumns,axis=1)\ny_train = df_train['stroke']\nX_train = np.asarray(X_train)\ny_train = np.asarray(y_train)\n\nX_test = df_test.drop(dropColumns,axis=1)\ny_test = df_test['stroke']\nX_test = np.asarray(X_test)\ny_test = np.asarray(y_test)\n\nprint(\"X_train.shape\", X_train.shape)\nprint(\"y_train.shape\", y_train.shape)\nprint(\"X_test.shape\", X_test.shape)\nprint(\"y_test.shape\", y_test.shape)",
   "execution_count": 8,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": "training with: 43400 rows\ntesting with: 10848\ntraining strokes: 783 no strokes: 42617\ntest strokes: 168 no strokes: 10680\nX_train.shape (3132, 20)\ny_train.shape (3132,)\nX_test.shape (720, 20)\ny_test.shape (720,)\n"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "def model_rank():\n    global network_history_rank\n    model = Sequential()\n    #mode.add(Embedding(1000, 64, input_length=10))\n    model.add(Dense(32,input_shape=(20,)))\n    model.add(Dropout(0.1))\n    model.add(Dense(32, kernel_initializer='normal'))\n    model.add(Dense(1, activation='sigmoid'))\n    \n    model.compile(loss='mean_squared_error', optimizer=Adam(lr=0.0001), metrics=['accuracy'])\n    \n    network_history_rank = model.fit(X_train_rank, y_train, batch_size=128, epochs=600, verbose=1)\n    return model",
   "execution_count": 9,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "\n\nX_train_rank = df_train[df_rank.index[1:21]]\nX_train_rank = np.asarray(X_train_rank)\n\nX_test_rank = df_test[df_rank.index[1:21]]\nX_test_rank = np.asarray(X_test_rank)\n\nprint(df_rank.index[1:21])\nprint(X_train_rank.shape)\nprint(y_train.shape)",
   "execution_count": 10,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": "Index(['age', 'heart_disease', 'avg_glucose_level', 'hypertension',\n       'ever_married', 'work_type_Self-employed', 'work_type_children',\n       'smoking_status_formerly smoked', 'smoking_status_unknown', 'bmi',\n       'gender_Male', 'gender_Female', 'work_type_Never_worked',\n       'smoking_status_smokes', 'work_type_Govt_job', 'work_type_Private',\n       'Residence_type_Rural', 'Residence_type_Urban', 'gender_Other',\n       'smoking_status_never smoked'],\n      dtype='object')\n(3132, 20)\n(3132,)\n"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "model = model()",
   "execution_count": 11,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": "Epoch 1/600\n3132/3132 [==============================] - 0s 46us/step - loss: 133.2390 - acc: 0.1791\nEpoch 2/600\n3132/3132 [==============================] - 0s 23us/step - loss: 4.1756 - acc: 0.6248\nEpoch 3/600\n3132/3132 [==============================] - 0s 23us/step - loss: 1.3473 - acc: 0.7069\nEpoch 4/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.8892 - acc: 0.7171\nEpoch 5/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.8327 - acc: 0.7245\nEpoch 6/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.5858 - acc: 0.7305\nEpoch 7/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.6940 - acc: 0.7299\nEpoch 8/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.6977 - acc: 0.7372\nEpoch 9/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.5130 - acc: 0.7382\nEpoch 10/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.5479 - acc: 0.7411\nEpoch 11/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.3793 - acc: 0.7395\nEpoch 12/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.4338 - acc: 0.7458\nEpoch 13/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.4288 - acc: 0.7411\nEpoch 14/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.3617 - acc: 0.7417\nEpoch 15/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2934 - acc: 0.7427\nEpoch 16/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.3446 - acc: 0.7414\nEpoch 17/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2933 - acc: 0.7452\nEpoch 18/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.3046 - acc: 0.7465\nEpoch 19/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2722 - acc: 0.7443\nEpoch 20/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2991 - acc: 0.7446\nEpoch 21/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2810 - acc: 0.7455\nEpoch 22/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2994 - acc: 0.7455\nEpoch 23/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2906 - acc: 0.7487\nEpoch 24/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2932 - acc: 0.7468\nEpoch 25/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2827 - acc: 0.7465\nEpoch 26/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2694 - acc: 0.7478\nEpoch 27/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.3017 - acc: 0.7465\nEpoch 28/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.3041 - acc: 0.7471\nEpoch 29/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2722 - acc: 0.7465\nEpoch 30/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.3042 - acc: 0.7471\nEpoch 31/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2827 - acc: 0.7474\nEpoch 32/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2631 - acc: 0.7468\nEpoch 33/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.4018 - acc: 0.7478\nEpoch 34/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2623 - acc: 0.7490\nEpoch 35/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2551 - acc: 0.7490\nEpoch 36/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2604 - acc: 0.7478\nEpoch 37/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2645 - acc: 0.7503\nEpoch 38/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.3231 - acc: 0.7481\nEpoch 39/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2522 - acc: 0.7497\nEpoch 40/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2604 - acc: 0.7484\nEpoch 41/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2590 - acc: 0.7481\nEpoch 42/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2953 - acc: 0.7474\nEpoch 43/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.3221 - acc: 0.7458\nEpoch 44/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2612 - acc: 0.7487\nEpoch 45/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2777 - acc: 0.7490\nEpoch 46/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2509 - acc: 0.7494\nEpoch 47/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2725 - acc: 0.7474\nEpoch 48/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2552 - acc: 0.7494\nEpoch 49/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2562 - acc: 0.7490\nEpoch 50/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2555 - acc: 0.7497\nEpoch 51/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2520 - acc: 0.7494\nEpoch 52/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2775 - acc: 0.7490\nEpoch 53/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2527 - acc: 0.7490\nEpoch 54/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2608 - acc: 0.7484\nEpoch 55/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2710 - acc: 0.7497\nEpoch 56/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2609 - acc: 0.7487\nEpoch 57/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2509 - acc: 0.7497\nEpoch 58/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2524 - acc: 0.7490\nEpoch 59/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2533 - acc: 0.7487\nEpoch 60/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2532 - acc: 0.7494\nEpoch 61/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2508 - acc: 0.7497\nEpoch 62/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2685 - acc: 0.7494\nEpoch 63/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2585 - acc: 0.7497\nEpoch 64/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2505 - acc: 0.7494\nEpoch 65/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2587 - acc: 0.7487\nEpoch 66/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2536 - acc: 0.7494\nEpoch 67/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2505 - acc: 0.7497\nEpoch 68/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2508 - acc: 0.7497\nEpoch 69/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2672 - acc: 0.7494\nEpoch 70/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2619 - acc: 0.7487\nEpoch 71/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2538 - acc: 0.7497\nEpoch 72/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2513 - acc: 0.7497\nEpoch 73/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2501 - acc: 0.7497\nEpoch 74/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2533 - acc: 0.7494\nEpoch 75/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2539 - acc: 0.7494\nEpoch 76/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2528 - acc: 0.7487\nEpoch 77/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2502 - acc: 0.7497\nEpoch 78/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2517 - acc: 0.7497\nEpoch 79/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2601 - acc: 0.7487\nEpoch 80/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2535 - acc: 0.7497\nEpoch 81/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2523 - acc: 0.7494\nEpoch 82/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2531 - acc: 0.7497\n"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": "Epoch 83/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2577 - acc: 0.7481\nEpoch 84/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2500 - acc: 0.7500\nEpoch 85/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2519 - acc: 0.7490\nEpoch 86/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 87/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 88/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2514 - acc: 0.7500\nEpoch 89/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2520 - acc: 0.7494\nEpoch 90/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2498 - acc: 0.7503\nEpoch 91/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2579 - acc: 0.7484\nEpoch 92/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2649 - acc: 0.7487\nEpoch 93/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2543 - acc: 0.7506\nEpoch 94/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2501 - acc: 0.7497\nEpoch 95/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2561 - acc: 0.7497\nEpoch 96/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2660 - acc: 0.7490\nEpoch 97/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2545 - acc: 0.7494\nEpoch 98/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2510 - acc: 0.7497\nEpoch 99/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2495 - acc: 0.7503\nEpoch 100/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 101/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2513 - acc: 0.7497\nEpoch 102/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2555 - acc: 0.7494\nEpoch 103/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 104/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2541 - acc: 0.7503\nEpoch 105/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 106/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2545 - acc: 0.7490\nEpoch 107/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2497 - acc: 0.7503\nEpoch 108/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 109/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2514 - acc: 0.7497\nEpoch 110/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2505 - acc: 0.7497\nEpoch 111/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2507 - acc: 0.7494\nEpoch 112/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2501 - acc: 0.7500\nEpoch 113/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2502 - acc: 0.7497\nEpoch 114/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2499 - acc: 0.7500\nEpoch 115/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2499 - acc: 0.7500\nEpoch 116/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2500 - acc: 0.7500\nEpoch 117/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2510 - acc: 0.7490\nEpoch 118/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2507 - acc: 0.7500\nEpoch 119/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2502 - acc: 0.7497\nEpoch 120/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2529 - acc: 0.7497\nEpoch 121/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2511 - acc: 0.7494\nEpoch 122/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2500 - acc: 0.7500\nEpoch 123/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2509 - acc: 0.7497\nEpoch 124/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2532 - acc: 0.7490\nEpoch 125/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2693 - acc: 0.7484\nEpoch 126/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 127/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2501 - acc: 0.7497\nEpoch 128/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 129/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 130/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2499 - acc: 0.7500\nEpoch 131/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 132/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2533 - acc: 0.7497\nEpoch 133/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2519 - acc: 0.7490\nEpoch 134/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2501 - acc: 0.7497\nEpoch 135/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2501 - acc: 0.7497\nEpoch 136/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2503 - acc: 0.7497\nEpoch 137/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2502 - acc: 0.7497\nEpoch 138/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2510 - acc: 0.7497\nEpoch 139/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2559 - acc: 0.7497\nEpoch 140/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2539 - acc: 0.7497\nEpoch 141/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2505 - acc: 0.7497\nEpoch 142/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 143/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2499 - acc: 0.7500\nEpoch 144/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 145/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2507 - acc: 0.7500\nEpoch 146/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 147/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 148/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2501 - acc: 0.7497\nEpoch 149/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 150/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2547 - acc: 0.7494\nEpoch 151/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 152/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2501 - acc: 0.7497\nEpoch 153/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 154/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 155/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2501 - acc: 0.7500\nEpoch 156/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 157/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2511 - acc: 0.7497\nEpoch 158/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 159/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2502 - acc: 0.7497\nEpoch 160/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2507 - acc: 0.7500\nEpoch 161/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2498 - acc: 0.7500\nEpoch 162/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2501 - acc: 0.7497\nEpoch 163/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 164/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2500 - acc: 0.7500\nEpoch 165/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2501 - acc: 0.7500\nEpoch 166/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2498 - acc: 0.7500\nEpoch 167/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2501 - acc: 0.7500\nEpoch 168/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2502 - acc: 0.7497\nEpoch 169/600\n3132/3132 [==============================] - 0s 21us/step - loss: 0.2498 - acc: 0.7500\nEpoch 170/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 171/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2501 - acc: 0.7500\nEpoch 172/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2500 - acc: 0.7500\nEpoch 173/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 174/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2502 - acc: 0.7497\nEpoch 175/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2498 - acc: 0.7500\nEpoch 176/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 177/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2505 - acc: 0.7497\nEpoch 178/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2623 - acc: 0.7494\nEpoch 179/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2501 - acc: 0.7500\nEpoch 180/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 181/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 182/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 183/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2508 - acc: 0.7497\nEpoch 184/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 185/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 186/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 187/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2508 - acc: 0.7497\nEpoch 188/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 189/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 190/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2500 - acc: 0.7500\nEpoch 191/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 192/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2497 - acc: 0.7503\nEpoch 193/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 194/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 195/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 196/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2577 - acc: 0.7494\nEpoch 197/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 198/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 199/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 200/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 201/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 202/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 203/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 204/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 205/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 206/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 207/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2497 - acc: 0.7503\nEpoch 208/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2514 - acc: 0.7497\nEpoch 209/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2497 - acc: 0.7503\nEpoch 210/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 211/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 212/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 213/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2497 - acc: 0.7503\nEpoch 214/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 215/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 216/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2547 - acc: 0.7500\nEpoch 217/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 218/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2507 - acc: 0.7497\nEpoch 219/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 220/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2501 - acc: 0.7500\nEpoch 221/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2501 - acc: 0.7500\nEpoch 222/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 223/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2499 - acc: 0.7500\nEpoch 224/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 225/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 226/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2544 - acc: 0.7497\nEpoch 227/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 228/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 229/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 230/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 231/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2533 - acc: 0.7497\nEpoch 232/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2500 - acc: 0.7500\nEpoch 233/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 234/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2501 - acc: 0.7500\nEpoch 235/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 236/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 237/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 238/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 239/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 240/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 241/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2501 - acc: 0.7500\nEpoch 242/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 243/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2527 - acc: 0.7497\nEpoch 244/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 245/600\n"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": "3132/3132 [==============================] - 0s 27us/step - loss: 0.2500 - acc: 0.7500\nEpoch 246/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.2500 - acc: 0.7500\nEpoch 247/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 248/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 249/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2500 - acc: 0.7500\nEpoch 250/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2515 - acc: 0.7500\nEpoch 251/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 252/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 253/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 254/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2506 - acc: 0.7497\nEpoch 255/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 256/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 257/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2500 - acc: 0.7500\nEpoch 258/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 259/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 260/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2497 - acc: 0.7503\nEpoch 261/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2503 - acc: 0.7497\nEpoch 262/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2498 - acc: 0.7503\nEpoch 263/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 264/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 265/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2500 - acc: 0.7500\nEpoch 266/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 267/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 268/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 269/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 270/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 271/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2504 - acc: 0.7497\nEpoch 272/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 273/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2501 - acc: 0.7497\nEpoch 274/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 275/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 276/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2519 - acc: 0.7497\nEpoch 277/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 278/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2508 - acc: 0.7497\nEpoch 279/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 280/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 281/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 282/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.2500 - acc: 0.7500\nEpoch 283/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 284/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 285/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2504 - acc: 0.7497\nEpoch 286/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 287/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 288/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 289/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 290/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 291/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 292/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 293/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 294/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 295/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 296/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 297/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2497 - acc: 0.7503\nEpoch 298/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 299/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2509 - acc: 0.7500\nEpoch 300/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 301/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.2500 - acc: 0.7500\nEpoch 302/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 303/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 304/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 305/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 306/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 307/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 308/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 309/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.2500 - acc: 0.7500\nEpoch 310/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 311/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2503 - acc: 0.7497\nEpoch 312/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 313/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 314/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2507 - acc: 0.7497\nEpoch 315/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 316/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2504 - acc: 0.7497\nEpoch 317/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 318/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 319/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 320/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 321/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 322/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 323/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 324/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 325/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 326/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 327/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.2500 - acc: 0.7500\nEpoch 328/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.2500 - acc: 0.7500\nEpoch 329/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 330/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 331/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 332/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 333/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 334/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 335/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 336/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 337/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2560 - acc: 0.7497\nEpoch 338/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 339/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 340/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 341/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 342/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 343/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 344/600\n3132/3132 [==============================] - 0s 21us/step - loss: 0.2498 - acc: 0.7500\nEpoch 345/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 346/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 347/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 348/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 349/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 350/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 351/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 352/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 353/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 354/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 355/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 356/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 357/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 358/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 359/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 360/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 361/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 362/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 363/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 364/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 365/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 366/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 367/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 368/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 369/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 370/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 371/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 372/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 373/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 374/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 375/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 376/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 377/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 378/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2497 - acc: 0.7503\nEpoch 379/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 380/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 381/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 382/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 383/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 384/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.2500 - acc: 0.7500\nEpoch 385/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 386/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 387/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 388/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2528 - acc: 0.7497\nEpoch 389/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 390/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 391/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 392/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 393/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 394/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 395/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 396/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 397/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 398/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 399/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 400/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 401/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 402/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 403/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 404/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 405/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 406/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 407/600\n"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": "3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 408/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 409/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 410/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 411/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 412/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 413/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 414/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 415/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 416/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 417/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 418/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 419/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 420/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 421/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 422/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 423/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2497 - acc: 0.7503\nEpoch 424/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 425/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 426/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 427/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 428/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 429/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 430/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 431/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 432/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 433/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 434/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 435/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 436/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 437/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 438/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 439/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2507 - acc: 0.7497\nEpoch 440/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 441/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 442/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 443/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 444/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 445/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 446/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 447/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 448/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 449/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 450/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 451/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 452/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 453/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 454/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 455/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 456/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 457/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 458/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 459/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 460/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 461/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 462/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 463/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 464/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 465/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 466/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 467/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 468/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 469/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 470/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 471/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 472/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 473/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 474/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 475/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.2500 - acc: 0.7500\nEpoch 476/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 477/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 478/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 479/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 480/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 481/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 482/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 483/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 484/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 485/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 486/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2501 - acc: 0.7497\nEpoch 487/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 488/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 489/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 490/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 491/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 492/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 493/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 494/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 495/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2500 - acc: 0.7500\nEpoch 496/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 497/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 498/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 499/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 500/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2500 - acc: 0.7500\nEpoch 501/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 502/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 503/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 504/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 505/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 506/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 507/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 508/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 509/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 510/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 511/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 512/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 513/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 514/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 515/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 516/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 517/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 518/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 519/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 520/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 521/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 522/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 523/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 524/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 525/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 526/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2500 - acc: 0.7500\nEpoch 527/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.2500 - acc: 0.7500\nEpoch 528/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 529/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 530/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 531/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 532/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 533/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 534/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 535/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 536/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 537/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 538/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 539/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 540/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 541/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 542/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 543/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 544/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 545/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 546/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 547/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 548/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 549/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 550/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 551/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 552/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.2500 - acc: 0.7500\nEpoch 553/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 554/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 555/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 556/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 557/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 558/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 559/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 560/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 561/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 562/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 563/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 564/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 565/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 566/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 567/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 568/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.2500 - acc: 0.7500\nEpoch 569/600\n"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": "3132/3132 [==============================] - 0s 27us/step - loss: 0.2500 - acc: 0.7500\nEpoch 570/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 571/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 572/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2500 - acc: 0.7500\nEpoch 573/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 574/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 575/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 576/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2500 - acc: 0.7500\nEpoch 577/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 578/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.2500 - acc: 0.7500\nEpoch 579/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.2500 - acc: 0.7500\nEpoch 580/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 581/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 582/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 583/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 584/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 585/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 586/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 587/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 588/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 589/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 590/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 591/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 592/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 593/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 594/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.2500 - acc: 0.7500\nEpoch 595/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 596/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 597/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 598/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2500 - acc: 0.7500\nEpoch 599/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\nEpoch 600/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2500 - acc: 0.7500\n"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "\nmodel.summary()",
   "execution_count": 12,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": "_________________________________________________________________\nLayer (type)                 Output Shape              Param #   \n=================================================================\ndense_1 (Dense)              (None, 64)                1344      \n_________________________________________________________________\ndropout_1 (Dropout)          (None, 64)                0         \n_________________________________________________________________\ndense_2 (Dense)              (None, 32)                2080      \n_________________________________________________________________\ndense_3 (Dense)              (None, 1)                 33        \n=================================================================\nTotal params: 3,457\nTrainable params: 3,457\nNon-trainable params: 0\n_________________________________________________________________\n"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "for layer in model.layers:\n    #pprint(dir(layer))\n    print(layer.name, layer.input_shape, layer.output_shape)",
   "execution_count": 13,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": "dense_1 (None, 20) (None, 64)\ndropout_1 (None, 64) (None, 64)\ndense_2 (None, 64) (None, 32)\ndense_3 (None, 32) (None, 1)\n"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "model_rank = model_rank()",
   "execution_count": 14,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": "Epoch 1/600\n3132/3132 [==============================] - 0s 58us/step - loss: 0.2481 - acc: 0.7449\nEpoch 2/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2254 - acc: 0.7363\nEpoch 3/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2203 - acc: 0.7302\nEpoch 4/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2150 - acc: 0.7318\nEpoch 5/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2049 - acc: 0.7411\nEpoch 6/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.2009 - acc: 0.7446\nEpoch 7/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1959 - acc: 0.7494\nEpoch 8/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.2019 - acc: 0.7356\nEpoch 9/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1951 - acc: 0.7401\nEpoch 10/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1819 - acc: 0.7538\nEpoch 11/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1916 - acc: 0.7423\nEpoch 12/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1842 - acc: 0.7535\nEpoch 13/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1814 - acc: 0.7554\nEpoch 14/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1743 - acc: 0.7567\nEpoch 15/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1831 - acc: 0.7532\nEpoch 16/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1865 - acc: 0.7478\nEpoch 17/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1733 - acc: 0.7503\nEpoch 18/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1709 - acc: 0.7554\nEpoch 19/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1702 - acc: 0.7637\nEpoch 20/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1714 - acc: 0.7593\nEpoch 21/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1694 - acc: 0.7580\nEpoch 22/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.1696 - acc: 0.7538\nEpoch 23/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1699 - acc: 0.7583\nEpoch 24/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1693 - acc: 0.7538\nEpoch 25/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1642 - acc: 0.7609\nEpoch 26/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1644 - acc: 0.7631\nEpoch 27/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1658 - acc: 0.7596\nEpoch 28/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.1653 - acc: 0.7634\nEpoch 29/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1646 - acc: 0.7621\nEpoch 30/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.1637 - acc: 0.7605\nEpoch 31/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.1596 - acc: 0.7625\nEpoch 32/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1594 - acc: 0.7692\nEpoch 33/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1666 - acc: 0.7494\nEpoch 34/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1606 - acc: 0.7609\nEpoch 35/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1595 - acc: 0.7663\nEpoch 36/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1624 - acc: 0.7583\nEpoch 37/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1602 - acc: 0.7586\nEpoch 38/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1596 - acc: 0.7660\nEpoch 39/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1614 - acc: 0.7637\nEpoch 40/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1607 - acc: 0.7640\nEpoch 41/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.1582 - acc: 0.7730\nEpoch 42/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1586 - acc: 0.7692\nEpoch 43/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1579 - acc: 0.7647\nEpoch 44/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1583 - acc: 0.7682\nEpoch 45/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1567 - acc: 0.7663\nEpoch 46/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.1562 - acc: 0.7663\nEpoch 47/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1586 - acc: 0.7656\nEpoch 48/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1568 - acc: 0.7676\nEpoch 49/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1585 - acc: 0.7679\nEpoch 50/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1564 - acc: 0.7685\nEpoch 51/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1560 - acc: 0.7733\nEpoch 52/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1587 - acc: 0.7698\nEpoch 53/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1604 - acc: 0.7609\nEpoch 54/600\n3132/3132 [==============================] - 0s 42us/step - loss: 0.1562 - acc: 0.7708\nEpoch 55/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1602 - acc: 0.7589\nEpoch 56/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1631 - acc: 0.7551\nEpoch 57/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1563 - acc: 0.7692\nEpoch 58/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1557 - acc: 0.7660\nEpoch 59/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1565 - acc: 0.7679\nEpoch 60/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1564 - acc: 0.7714\nEpoch 61/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1540 - acc: 0.7682\nEpoch 62/600\n3132/3132 [==============================] - 0s 31us/step - loss: 0.1551 - acc: 0.7723\nEpoch 63/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1540 - acc: 0.7692\nEpoch 64/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1552 - acc: 0.7708\nEpoch 65/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1538 - acc: 0.7653\nEpoch 66/600\n3132/3132 [==============================] - 0s 32us/step - loss: 0.1525 - acc: 0.7762\nEpoch 67/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1526 - acc: 0.7688\nEpoch 68/600\n3132/3132 [==============================] - 0s 31us/step - loss: 0.1530 - acc: 0.7688\nEpoch 69/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1557 - acc: 0.7720\nEpoch 70/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1554 - acc: 0.7663\nEpoch 71/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1537 - acc: 0.7714\nEpoch 72/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1543 - acc: 0.7768\nEpoch 73/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1537 - acc: 0.7711\nEpoch 74/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1532 - acc: 0.7739\nEpoch 75/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1540 - acc: 0.7688\nEpoch 76/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1532 - acc: 0.7714\nEpoch 77/600\n3132/3132 [==============================] - 0s 31us/step - loss: 0.1551 - acc: 0.7644\nEpoch 78/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1516 - acc: 0.7765\nEpoch 79/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1525 - acc: 0.7797\nEpoch 80/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1522 - acc: 0.7666\nEpoch 81/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1527 - acc: 0.7727\nEpoch 82/600\n3132/3132 [==============================] - 0s 32us/step - loss: 0.1508 - acc: 0.7778\nEpoch 83/600\n"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": "3132/3132 [==============================] - 0s 32us/step - loss: 0.1533 - acc: 0.7711\nEpoch 84/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1528 - acc: 0.7650\nEpoch 85/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1508 - acc: 0.7714\nEpoch 86/600\n3132/3132 [==============================] - 0s 30us/step - loss: 0.1516 - acc: 0.7733\nEpoch 87/600\n3132/3132 [==============================] - 0s 32us/step - loss: 0.1518 - acc: 0.7708\nEpoch 88/600\n3132/3132 [==============================] - 0s 32us/step - loss: 0.1493 - acc: 0.7768\nEpoch 89/600\n3132/3132 [==============================] - 0s 30us/step - loss: 0.1505 - acc: 0.7755\nEpoch 90/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1535 - acc: 0.7739\nEpoch 91/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1513 - acc: 0.7736\nEpoch 92/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1500 - acc: 0.7746\nEpoch 93/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.1520 - acc: 0.7746\nEpoch 94/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1497 - acc: 0.7800\nEpoch 95/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1480 - acc: 0.7848\nEpoch 96/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1540 - acc: 0.7672\nEpoch 97/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1490 - acc: 0.7727\nEpoch 98/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1499 - acc: 0.7749\nEpoch 99/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1514 - acc: 0.7717\nEpoch 100/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1489 - acc: 0.7695\nEpoch 101/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1488 - acc: 0.7807\nEpoch 102/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1491 - acc: 0.7755\nEpoch 103/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.1486 - acc: 0.7787\nEpoch 104/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1493 - acc: 0.7765\nEpoch 105/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1476 - acc: 0.7752\nEpoch 106/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1477 - acc: 0.7819\nEpoch 107/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1468 - acc: 0.7842\nEpoch 108/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1483 - acc: 0.7768\nEpoch 109/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1483 - acc: 0.7730\nEpoch 110/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1494 - acc: 0.7688\nEpoch 111/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1468 - acc: 0.7848\nEpoch 112/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1503 - acc: 0.7749\nEpoch 113/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1475 - acc: 0.7800\nEpoch 114/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1475 - acc: 0.7797\nEpoch 115/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1471 - acc: 0.7791\nEpoch 116/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1489 - acc: 0.7778\nEpoch 117/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1460 - acc: 0.7832\nEpoch 118/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1458 - acc: 0.7854\nEpoch 119/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1478 - acc: 0.7771\nEpoch 120/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1463 - acc: 0.7822\nEpoch 121/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1486 - acc: 0.7775\nEpoch 122/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1470 - acc: 0.7813\nEpoch 123/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1467 - acc: 0.7851\nEpoch 124/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1457 - acc: 0.7816\nEpoch 125/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1459 - acc: 0.7816\nEpoch 126/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1460 - acc: 0.7768\nEpoch 127/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1471 - acc: 0.7768\nEpoch 128/600\n3132/3132 [==============================] - 0s 30us/step - loss: 0.1483 - acc: 0.7832\nEpoch 129/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1451 - acc: 0.7819\nEpoch 130/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1469 - acc: 0.7749\nEpoch 131/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1454 - acc: 0.7822\nEpoch 132/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1462 - acc: 0.7781\nEpoch 133/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1462 - acc: 0.7759\nEpoch 134/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1457 - acc: 0.7810\nEpoch 135/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1448 - acc: 0.7765\nEpoch 136/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1439 - acc: 0.7842\nEpoch 137/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1440 - acc: 0.7864\nEpoch 138/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1456 - acc: 0.7832\nEpoch 139/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1468 - acc: 0.7778\nEpoch 140/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1476 - acc: 0.7714\nEpoch 141/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1435 - acc: 0.7835\nEpoch 142/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1461 - acc: 0.7822\nEpoch 143/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1455 - acc: 0.7778\nEpoch 144/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1441 - acc: 0.7781\nEpoch 145/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1467 - acc: 0.7775\nEpoch 146/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1424 - acc: 0.7835\nEpoch 147/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1449 - acc: 0.7813\nEpoch 148/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1439 - acc: 0.7874\nEpoch 149/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1440 - acc: 0.7826\nEpoch 150/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1448 - acc: 0.7870\nEpoch 151/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1443 - acc: 0.7835\nEpoch 152/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1444 - acc: 0.7822\nEpoch 153/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1432 - acc: 0.7819\nEpoch 154/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1436 - acc: 0.7861\nEpoch 155/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1446 - acc: 0.7842\nEpoch 156/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1440 - acc: 0.7813\nEpoch 157/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1450 - acc: 0.7816\nEpoch 158/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1446 - acc: 0.7819\nEpoch 159/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1445 - acc: 0.7832\nEpoch 160/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1447 - acc: 0.7842\nEpoch 161/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1477 - acc: 0.7781\nEpoch 162/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1449 - acc: 0.7784\nEpoch 163/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1431 - acc: 0.7883\nEpoch 164/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1433 - acc: 0.7845\nEpoch 165/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1450 - acc: 0.7794\nEpoch 166/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1430 - acc: 0.7813\nEpoch 167/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1426 - acc: 0.7867\nEpoch 168/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1433 - acc: 0.7874\nEpoch 169/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1435 - acc: 0.7819\nEpoch 170/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1445 - acc: 0.7819\nEpoch 171/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1435 - acc: 0.7842\nEpoch 172/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1425 - acc: 0.7883\nEpoch 173/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1433 - acc: 0.7867\nEpoch 174/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1424 - acc: 0.7902\nEpoch 175/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1428 - acc: 0.7893\nEpoch 176/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.1441 - acc: 0.7816\nEpoch 177/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.1435 - acc: 0.7816\nEpoch 178/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1433 - acc: 0.7822\nEpoch 179/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1427 - acc: 0.7848\nEpoch 180/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1437 - acc: 0.7826\nEpoch 181/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1433 - acc: 0.7902\nEpoch 182/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1423 - acc: 0.7870\nEpoch 183/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1415 - acc: 0.7854\nEpoch 184/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1426 - acc: 0.7813\nEpoch 185/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1430 - acc: 0.7877\nEpoch 186/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1414 - acc: 0.7909\nEpoch 187/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1416 - acc: 0.7893\nEpoch 188/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1439 - acc: 0.7822\nEpoch 189/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1422 - acc: 0.7880\nEpoch 190/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1423 - acc: 0.7874\nEpoch 191/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1415 - acc: 0.7838\nEpoch 192/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1428 - acc: 0.7832\nEpoch 193/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1440 - acc: 0.7854\nEpoch 194/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1441 - acc: 0.7829\nEpoch 195/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1448 - acc: 0.7822\nEpoch 196/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1418 - acc: 0.7870\nEpoch 197/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1426 - acc: 0.7829\nEpoch 198/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1416 - acc: 0.7890\nEpoch 199/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1418 - acc: 0.7845\nEpoch 200/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1415 - acc: 0.7842\nEpoch 201/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1423 - acc: 0.7867\nEpoch 202/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1424 - acc: 0.7838\nEpoch 203/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1408 - acc: 0.7864\nEpoch 204/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1417 - acc: 0.7896\nEpoch 205/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1408 - acc: 0.7937\nEpoch 206/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1438 - acc: 0.7813\nEpoch 207/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1415 - acc: 0.7867\nEpoch 208/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1424 - acc: 0.7835\nEpoch 209/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1412 - acc: 0.7867\nEpoch 210/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1421 - acc: 0.7838\nEpoch 211/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1415 - acc: 0.7829\nEpoch 212/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1410 - acc: 0.7874\nEpoch 213/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1417 - acc: 0.7934\nEpoch 214/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1437 - acc: 0.7835\nEpoch 215/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1425 - acc: 0.7880\nEpoch 216/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1412 - acc: 0.7874\nEpoch 217/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1400 - acc: 0.7909\nEpoch 218/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1417 - acc: 0.7838\nEpoch 219/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1409 - acc: 0.7921\nEpoch 220/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1395 - acc: 0.7902\nEpoch 221/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1419 - acc: 0.7867\nEpoch 222/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1401 - acc: 0.7928\nEpoch 223/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1406 - acc: 0.7845\nEpoch 224/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1402 - acc: 0.7883\nEpoch 225/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1436 - acc: 0.7874\nEpoch 226/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1411 - acc: 0.7905\nEpoch 227/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1406 - acc: 0.7925\nEpoch 228/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1402 - acc: 0.7909\nEpoch 229/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1410 - acc: 0.7918\nEpoch 230/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1395 - acc: 0.7886\nEpoch 231/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1398 - acc: 0.7950\nEpoch 232/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1414 - acc: 0.7861\nEpoch 233/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1413 - acc: 0.7867\nEpoch 234/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1404 - acc: 0.7890\nEpoch 235/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1405 - acc: 0.7864\nEpoch 236/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1395 - acc: 0.7915\nEpoch 237/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1408 - acc: 0.7893\nEpoch 238/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1397 - acc: 0.7918\nEpoch 239/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1418 - acc: 0.7851\nEpoch 240/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1450 - acc: 0.7810\nEpoch 241/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1414 - acc: 0.7854\nEpoch 242/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1386 - acc: 0.7941\nEpoch 243/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1393 - acc: 0.7912\nEpoch 244/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1395 - acc: 0.7931\nEpoch 245/600\n"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": "3132/3132 [==============================] - 0s 27us/step - loss: 0.1400 - acc: 0.7858\nEpoch 246/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1397 - acc: 0.7921\nEpoch 247/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1421 - acc: 0.7858\nEpoch 248/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1389 - acc: 0.7912\nEpoch 249/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.1411 - acc: 0.7835\nEpoch 250/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1408 - acc: 0.7921\nEpoch 251/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1406 - acc: 0.7880\nEpoch 252/600\n3132/3132 [==============================] - 0s 30us/step - loss: 0.1415 - acc: 0.7880\nEpoch 253/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1400 - acc: 0.7953\nEpoch 254/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1404 - acc: 0.7921\nEpoch 255/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1416 - acc: 0.7890\nEpoch 256/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1410 - acc: 0.7909\nEpoch 257/600\n3132/3132 [==============================] - 0s 40us/step - loss: 0.1386 - acc: 0.7963\nEpoch 258/600\n3132/3132 [==============================] - 0s 32us/step - loss: 0.1401 - acc: 0.7854\nEpoch 259/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1406 - acc: 0.7864\nEpoch 260/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.1394 - acc: 0.7934\nEpoch 261/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.1403 - acc: 0.7931\nEpoch 262/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1410 - acc: 0.7874\nEpoch 263/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.1412 - acc: 0.7858\nEpoch 264/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1391 - acc: 0.7883\nEpoch 265/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1398 - acc: 0.7883\nEpoch 266/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1389 - acc: 0.7921\nEpoch 267/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1392 - acc: 0.7918\nEpoch 268/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1418 - acc: 0.7867\nEpoch 269/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1390 - acc: 0.7960\nEpoch 270/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1374 - acc: 0.7969\nEpoch 271/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1398 - acc: 0.7918\nEpoch 272/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1380 - acc: 0.7973\nEpoch 273/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1395 - acc: 0.7893\nEpoch 274/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1396 - acc: 0.7899\nEpoch 275/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.1392 - acc: 0.7886\nEpoch 276/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1389 - acc: 0.7893\nEpoch 277/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1383 - acc: 0.7982\nEpoch 278/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1395 - acc: 0.7909\nEpoch 279/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.1400 - acc: 0.7896\nEpoch 280/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1388 - acc: 0.7925\nEpoch 281/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1403 - acc: 0.7896\nEpoch 282/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1386 - acc: 0.7890\nEpoch 283/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1386 - acc: 0.7896\nEpoch 284/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1395 - acc: 0.7870\nEpoch 285/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1398 - acc: 0.7912\nEpoch 286/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1378 - acc: 0.7899\nEpoch 287/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1403 - acc: 0.7861\nEpoch 288/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1403 - acc: 0.7880\nEpoch 289/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1392 - acc: 0.7877\nEpoch 290/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1399 - acc: 0.7858\nEpoch 291/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1392 - acc: 0.7925\nEpoch 292/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.1408 - acc: 0.7899\nEpoch 293/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1380 - acc: 0.7928\nEpoch 294/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1394 - acc: 0.7877\nEpoch 295/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1386 - acc: 0.7915\nEpoch 296/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1383 - acc: 0.7921\nEpoch 297/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1389 - acc: 0.7912\nEpoch 298/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1388 - acc: 0.7957\nEpoch 299/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1388 - acc: 0.7921\nEpoch 300/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1378 - acc: 0.7902\nEpoch 301/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1408 - acc: 0.7890\nEpoch 302/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1385 - acc: 0.7931\nEpoch 303/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1384 - acc: 0.7909\nEpoch 304/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1398 - acc: 0.7870\nEpoch 305/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1405 - acc: 0.7851\nEpoch 306/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1382 - acc: 0.7915\nEpoch 307/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1388 - acc: 0.7973\nEpoch 308/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1393 - acc: 0.7915\nEpoch 309/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1393 - acc: 0.7886\nEpoch 310/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1387 - acc: 0.7950\nEpoch 311/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1390 - acc: 0.7893\nEpoch 312/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1380 - acc: 0.7909\nEpoch 313/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1377 - acc: 0.7928\nEpoch 314/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1376 - acc: 0.7918\nEpoch 315/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1365 - acc: 0.7976\nEpoch 316/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1380 - acc: 0.7931\nEpoch 317/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1383 - acc: 0.7902\nEpoch 318/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1397 - acc: 0.7915\nEpoch 319/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1382 - acc: 0.7899\nEpoch 320/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1397 - acc: 0.7883\nEpoch 321/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1407 - acc: 0.7867\nEpoch 322/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1388 - acc: 0.7838\nEpoch 323/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1390 - acc: 0.7890\nEpoch 324/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1393 - acc: 0.7864\nEpoch 325/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1412 - acc: 0.7893\nEpoch 326/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1383 - acc: 0.7928\nEpoch 327/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1379 - acc: 0.7890\nEpoch 328/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1384 - acc: 0.7934\nEpoch 329/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1383 - acc: 0.7921\nEpoch 330/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1392 - acc: 0.7944\nEpoch 331/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1374 - acc: 0.7941\nEpoch 332/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1381 - acc: 0.7864\nEpoch 333/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1381 - acc: 0.7947\nEpoch 334/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1383 - acc: 0.7893\nEpoch 335/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1367 - acc: 0.7995\nEpoch 336/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1381 - acc: 0.7918\nEpoch 337/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1389 - acc: 0.7928\nEpoch 338/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1387 - acc: 0.7880\nEpoch 339/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1381 - acc: 0.7893\nEpoch 340/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1369 - acc: 0.7912\nEpoch 341/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1366 - acc: 0.7957\nEpoch 342/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1369 - acc: 0.7937\nEpoch 343/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1370 - acc: 0.7966\nEpoch 344/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1374 - acc: 0.7941\nEpoch 345/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1384 - acc: 0.7861\nEpoch 346/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1372 - acc: 0.7969\nEpoch 347/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1368 - acc: 0.7928\nEpoch 348/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1374 - acc: 0.7985\nEpoch 349/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1380 - acc: 0.7934\nEpoch 350/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1387 - acc: 0.7883\nEpoch 351/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1372 - acc: 0.7918\nEpoch 352/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1374 - acc: 0.7918\nEpoch 353/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1367 - acc: 0.8030\nEpoch 354/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1376 - acc: 0.7925\nEpoch 355/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1376 - acc: 0.7928\nEpoch 356/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1392 - acc: 0.7886\nEpoch 357/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1385 - acc: 0.7950\nEpoch 358/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1387 - acc: 0.7912\nEpoch 359/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1376 - acc: 0.7960\nEpoch 360/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1372 - acc: 0.7893\nEpoch 361/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1376 - acc: 0.7960\nEpoch 362/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1382 - acc: 0.7902\nEpoch 363/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1381 - acc: 0.7925\nEpoch 364/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1371 - acc: 0.7877\nEpoch 365/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1377 - acc: 0.7947\nEpoch 366/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1387 - acc: 0.7941\nEpoch 367/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1369 - acc: 0.7931\nEpoch 368/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1378 - acc: 0.7905\nEpoch 369/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1373 - acc: 0.7934\nEpoch 370/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1379 - acc: 0.7928\nEpoch 371/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1375 - acc: 0.7918\nEpoch 372/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1380 - acc: 0.7912\nEpoch 373/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1376 - acc: 0.7909\nEpoch 374/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1385 - acc: 0.7886\nEpoch 375/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1388 - acc: 0.7880\nEpoch 376/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1372 - acc: 0.7950\nEpoch 377/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1370 - acc: 0.7976\nEpoch 378/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1381 - acc: 0.7944\nEpoch 379/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1382 - acc: 0.7934\nEpoch 380/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1376 - acc: 0.7937\nEpoch 381/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1377 - acc: 0.7899\nEpoch 382/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1391 - acc: 0.7870\nEpoch 383/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1381 - acc: 0.7989\nEpoch 384/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1371 - acc: 0.7921\nEpoch 385/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1364 - acc: 0.7931\nEpoch 386/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1378 - acc: 0.7963\nEpoch 387/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1379 - acc: 0.7947\nEpoch 388/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1391 - acc: 0.7905\nEpoch 389/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1376 - acc: 0.7941\nEpoch 390/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1367 - acc: 0.7941\nEpoch 391/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1383 - acc: 0.7921\nEpoch 392/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1390 - acc: 0.7934\nEpoch 393/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1375 - acc: 0.7928\nEpoch 394/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1386 - acc: 0.7893\nEpoch 395/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1402 - acc: 0.7896\nEpoch 396/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1372 - acc: 0.7944\nEpoch 397/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1374 - acc: 0.7899\nEpoch 398/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1370 - acc: 0.7934\nEpoch 399/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1358 - acc: 0.7969\nEpoch 400/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1363 - acc: 0.7918\nEpoch 401/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1364 - acc: 0.7925\nEpoch 402/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1363 - acc: 0.7886\nEpoch 403/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1369 - acc: 0.7931\nEpoch 404/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1364 - acc: 0.8033\nEpoch 405/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1383 - acc: 0.7915\nEpoch 406/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1376 - acc: 0.7899\nEpoch 407/600\n"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": "3132/3132 [==============================] - 0s 29us/step - loss: 0.1368 - acc: 0.7937\nEpoch 408/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1370 - acc: 0.7905\nEpoch 409/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1363 - acc: 0.7976\nEpoch 410/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1360 - acc: 0.7950\nEpoch 411/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1367 - acc: 0.7944\nEpoch 412/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1361 - acc: 0.7963\nEpoch 413/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1367 - acc: 0.7953\nEpoch 414/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1374 - acc: 0.7963\nEpoch 415/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1369 - acc: 0.7909\nEpoch 416/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1368 - acc: 0.8011\nEpoch 417/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1360 - acc: 0.7947\nEpoch 418/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1370 - acc: 0.8008\nEpoch 419/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1367 - acc: 0.7937\nEpoch 420/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1363 - acc: 0.7902\nEpoch 421/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1372 - acc: 0.7918\nEpoch 422/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.1370 - acc: 0.7960\nEpoch 423/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1367 - acc: 0.7973\nEpoch 424/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1364 - acc: 0.7944\nEpoch 425/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1363 - acc: 0.8001\nEpoch 426/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1370 - acc: 0.7973\nEpoch 427/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.1364 - acc: 0.7941\nEpoch 428/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1360 - acc: 0.8008\nEpoch 429/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1369 - acc: 0.7953\nEpoch 430/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1358 - acc: 0.7941\nEpoch 431/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1361 - acc: 0.7931\nEpoch 432/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.1373 - acc: 0.7928\nEpoch 433/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1370 - acc: 0.7957\nEpoch 434/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.1357 - acc: 0.7960\nEpoch 435/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1361 - acc: 0.7969\nEpoch 436/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1379 - acc: 0.7902\nEpoch 437/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1354 - acc: 0.7960\nEpoch 438/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1377 - acc: 0.7960\nEpoch 439/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1380 - acc: 0.7896\nEpoch 440/600\n3132/3132 [==============================] - 0s 22us/step - loss: 0.1365 - acc: 0.7886\nEpoch 441/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1376 - acc: 0.7877\nEpoch 442/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.1365 - acc: 0.7928\nEpoch 443/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1368 - acc: 0.7915\nEpoch 444/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1356 - acc: 0.7979\nEpoch 445/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1386 - acc: 0.7912\nEpoch 446/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1362 - acc: 0.7883\nEpoch 447/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1364 - acc: 0.7969\nEpoch 448/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1357 - acc: 0.7953\nEpoch 449/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1366 - acc: 0.8008\nEpoch 450/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1367 - acc: 0.7950\nEpoch 451/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1376 - acc: 0.7912\nEpoch 452/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1369 - acc: 0.7912\nEpoch 453/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1365 - acc: 0.7918\nEpoch 454/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1367 - acc: 0.7931\nEpoch 455/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1371 - acc: 0.8008\nEpoch 456/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1363 - acc: 0.7966\nEpoch 457/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1350 - acc: 0.7960\nEpoch 458/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1371 - acc: 0.7905\nEpoch 459/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1363 - acc: 0.7992\nEpoch 460/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1355 - acc: 0.7950\nEpoch 461/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1366 - acc: 0.7899\nEpoch 462/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1379 - acc: 0.7918\nEpoch 463/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1361 - acc: 0.7947\nEpoch 464/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1370 - acc: 0.7886\nEpoch 465/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1371 - acc: 0.7915\nEpoch 466/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1371 - acc: 0.7931\nEpoch 467/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1365 - acc: 0.7969\nEpoch 468/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1364 - acc: 0.7931\nEpoch 469/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1357 - acc: 0.7915\nEpoch 470/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1369 - acc: 0.7921\nEpoch 471/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1372 - acc: 0.7925\nEpoch 472/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1371 - acc: 0.7957\nEpoch 473/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1378 - acc: 0.7912\nEpoch 474/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1358 - acc: 0.7921\nEpoch 475/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1351 - acc: 0.7947\nEpoch 476/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1358 - acc: 0.7979\nEpoch 477/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1355 - acc: 0.7937\nEpoch 478/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1358 - acc: 0.7973\nEpoch 479/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1358 - acc: 0.7953\nEpoch 480/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1367 - acc: 0.7963\nEpoch 481/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1365 - acc: 0.7928\nEpoch 482/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1351 - acc: 0.7998\nEpoch 483/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1361 - acc: 0.7912\nEpoch 484/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1352 - acc: 0.7989\nEpoch 485/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1364 - acc: 0.7931\nEpoch 486/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1358 - acc: 0.7937\nEpoch 487/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1361 - acc: 0.7957\nEpoch 488/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1368 - acc: 0.7947\nEpoch 489/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1365 - acc: 0.7973\nEpoch 490/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1358 - acc: 0.7934\nEpoch 491/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1355 - acc: 0.8030\nEpoch 492/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1370 - acc: 0.7877\nEpoch 493/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1379 - acc: 0.7902\nEpoch 494/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1365 - acc: 0.7925\nEpoch 495/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1355 - acc: 0.7937\nEpoch 496/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1366 - acc: 0.7960\nEpoch 497/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1360 - acc: 0.8004\nEpoch 498/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1351 - acc: 0.7969\nEpoch 499/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1358 - acc: 0.7960\nEpoch 500/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1367 - acc: 0.7921\nEpoch 501/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1352 - acc: 0.7944\nEpoch 502/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1363 - acc: 0.7985\nEpoch 503/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1362 - acc: 0.7905\nEpoch 504/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1349 - acc: 0.7979\nEpoch 505/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1363 - acc: 0.7995\nEpoch 506/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1361 - acc: 0.7953\nEpoch 507/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1360 - acc: 0.7973\nEpoch 508/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1361 - acc: 0.7989\nEpoch 509/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1373 - acc: 0.7960\nEpoch 510/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1362 - acc: 0.7944\nEpoch 511/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1367 - acc: 0.7947\nEpoch 512/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1356 - acc: 0.7985\nEpoch 513/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1351 - acc: 0.8001\nEpoch 514/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1366 - acc: 0.7947\nEpoch 515/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1358 - acc: 0.7947\nEpoch 516/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1370 - acc: 0.7915\nEpoch 517/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1352 - acc: 0.7953\nEpoch 518/600\n3132/3132 [==============================] - 0s 23us/step - loss: 0.1359 - acc: 0.7957\nEpoch 519/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1360 - acc: 0.8017\nEpoch 520/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1376 - acc: 0.7909\nEpoch 521/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1360 - acc: 0.7944\nEpoch 522/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1366 - acc: 0.7947\nEpoch 523/600\n3132/3132 [==============================] - 0s 31us/step - loss: 0.1357 - acc: 0.7966\nEpoch 524/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1349 - acc: 0.7989\nEpoch 525/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1361 - acc: 0.7941\nEpoch 526/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1360 - acc: 0.7957\nEpoch 527/600\n3132/3132 [==============================] - 0s 32us/step - loss: 0.1361 - acc: 0.7973\nEpoch 528/600\n3132/3132 [==============================] - 0s 31us/step - loss: 0.1361 - acc: 0.7944\nEpoch 529/600\n3132/3132 [==============================] - 0s 30us/step - loss: 0.1354 - acc: 0.7976\nEpoch 530/600\n3132/3132 [==============================] - 0s 31us/step - loss: 0.1357 - acc: 0.7966\nEpoch 531/600\n3132/3132 [==============================] - 0s 30us/step - loss: 0.1361 - acc: 0.7941\nEpoch 532/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1359 - acc: 0.8027\nEpoch 533/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1364 - acc: 0.7937\nEpoch 534/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1350 - acc: 0.7944\nEpoch 535/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1368 - acc: 0.7851\nEpoch 536/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1359 - acc: 0.7982\nEpoch 537/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1352 - acc: 0.7976\nEpoch 538/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1356 - acc: 0.7973\nEpoch 539/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1355 - acc: 0.7925\nEpoch 540/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1352 - acc: 0.7976\nEpoch 541/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1357 - acc: 0.7957\nEpoch 542/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1355 - acc: 0.8017\nEpoch 543/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1361 - acc: 0.7953\nEpoch 544/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1346 - acc: 0.8033\nEpoch 545/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1360 - acc: 0.7953\nEpoch 546/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1355 - acc: 0.7960\nEpoch 547/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1356 - acc: 0.7899\nEpoch 548/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1362 - acc: 0.7905\nEpoch 549/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1359 - acc: 0.7953\nEpoch 550/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1360 - acc: 0.7937\nEpoch 551/600\n3132/3132 [==============================] - 0s 29us/step - loss: 0.1372 - acc: 0.7982\nEpoch 552/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1373 - acc: 0.7918\nEpoch 553/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1356 - acc: 0.7941\nEpoch 554/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1348 - acc: 0.7947\nEpoch 555/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1369 - acc: 0.7925\nEpoch 556/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1369 - acc: 0.7957\nEpoch 557/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1356 - acc: 0.7960\nEpoch 558/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1358 - acc: 0.7960\nEpoch 559/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1359 - acc: 0.7960\nEpoch 560/600\n3132/3132 [==============================] - 0s 28us/step - loss: 0.1357 - acc: 0.7976\nEpoch 561/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1358 - acc: 0.7925\nEpoch 562/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1346 - acc: 0.7979\nEpoch 563/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1359 - acc: 0.7950\nEpoch 564/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1358 - acc: 0.7960\nEpoch 565/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1338 - acc: 0.8030\nEpoch 566/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1355 - acc: 0.7941\nEpoch 567/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1354 - acc: 0.7973\nEpoch 568/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1359 - acc: 0.7915\nEpoch 569/600\n"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": "3132/3132 [==============================] - 0s 25us/step - loss: 0.1351 - acc: 0.7957\nEpoch 570/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1356 - acc: 0.7918\nEpoch 571/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1351 - acc: 0.7966\nEpoch 572/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1350 - acc: 0.7979\nEpoch 573/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1357 - acc: 0.7982\nEpoch 574/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1346 - acc: 0.8040\nEpoch 575/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1365 - acc: 0.7976\nEpoch 576/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1348 - acc: 0.7973\nEpoch 577/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1359 - acc: 0.7880\nEpoch 578/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1369 - acc: 0.7905\nEpoch 579/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1342 - acc: 0.8036\nEpoch 580/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1371 - acc: 0.7953\nEpoch 581/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1364 - acc: 0.7973\nEpoch 582/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1360 - acc: 0.7995\nEpoch 583/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1347 - acc: 0.7934\nEpoch 584/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1345 - acc: 0.7931\nEpoch 585/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1366 - acc: 0.7950\nEpoch 586/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1351 - acc: 0.7976\nEpoch 587/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1350 - acc: 0.7921\nEpoch 588/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1362 - acc: 0.7928\nEpoch 589/600\n3132/3132 [==============================] - 0s 27us/step - loss: 0.1352 - acc: 0.7950\nEpoch 590/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1351 - acc: 0.7973\nEpoch 591/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1351 - acc: 0.8020\nEpoch 592/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1346 - acc: 0.7969\nEpoch 593/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1355 - acc: 0.7957\nEpoch 594/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1348 - acc: 0.8040\nEpoch 595/600\n3132/3132 [==============================] - 0s 24us/step - loss: 0.1348 - acc: 0.8014\nEpoch 596/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1343 - acc: 0.7973\nEpoch 597/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1350 - acc: 0.7969\nEpoch 598/600\n3132/3132 [==============================] - 0s 26us/step - loss: 0.1369 - acc: 0.7899\nEpoch 599/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1343 - acc: 0.7992\nEpoch 600/600\n3132/3132 [==============================] - 0s 25us/step - loss: 0.1353 - acc: 0.7957\n"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "def plot_history(network_history,title='Loss and accuracy (Keras model)'):\n    plt.figure(figsize=(15,10))\n    plt.subplot(211)\n    plt.title(title)\n    plt.xlabel('Epochs')\n    plt.ylabel('Loss')\n    plt.plot(network_history.history['loss'])\n    #plt.plot(network_history.history['val_loss'])\n    plt.legend(['Training', 'Validation'])\n\n    plt.subplot(212)\n    plt.xlabel('Epochs')\n    plt.ylabel('Accuracy')\n    plt.plot(network_history.history['acc'])\n    #plt.plot(network_history.history['val_acc'])\n    plt.legend(['Training', 'Validation'], loc='lower right')\n    plt.show()",
   "execution_count": 15,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "plot_history(network_history)\n",
   "execution_count": 16,
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": "<matplotlib.figure.Figure at 0x1a3027a940>"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "plot_history(network_history_rank, 'Loss and accuracy with top 10 ranked (Keras model)')\n",
   "execution_count": 17,
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": "<matplotlib.figure.Figure at 0x109c90b70>"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "model.evaluate(X_test, y_test)",
   "execution_count": 18,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": "720/720 [==============================] - 0s 64us/step\n"
    },
    {
     "data": {
      "text/plain": "[0.23333333333333334, 0.76666666666666672]"
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "model_rank.evaluate(X_test_rank,y_test)",
   "execution_count": 19,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": "720/720 [==============================] - 0s 65us/step\n"
    },
    {
     "data": {
      "text/plain": "[0.12573329309622447, 0.82222222222222219]"
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "prediction = model.predict(X_test)",
   "execution_count": 20,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "prediction_rank = model_rank.predict(X_test_rank)",
   "execution_count": 21,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "plt.figure(figsize=(15,8))\nplt.subplot(121)\nplt.title('Histogram prediction')\nplt.hist(prediction)\n\nplt.subplot(122)\nplt.title('Histogram prediction rank')\nplt.hist(prediction_rank)\nplt.show()",
   "execution_count": 22,
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": "<matplotlib.figure.Figure at 0x1a3053f940>"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "prediction_bin = prediction > 0.5\nprediction_bin = prediction_bin.astype(int)\nconfusion_matrix(y_test, prediction_bin)",
   "execution_count": 23,
   "outputs": [
    {
     "data": {
      "text/plain": "array([[552,   0],\n       [168,   0]])"
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "prediction_rank_bin = prediction_rank > 0.5\nprediction_rank_bin = prediction_rank_bin.astype(int)\nconfusion_matrix(y_test, prediction_rank_bin)",
   "execution_count": 24,
   "outputs": [
    {
     "data": {
      "text/plain": "array([[493,  59],\n       [ 69,  99]])"
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "df_test_0_only = df_test_0.sample(200)\nX_test_0 = df_test_0_only.drop(dropColumns, axis=1)\nX_test_0_rank = df_test_0_only[df_rank.index[1:21]]\ny_test_0 = df_test_0_only['stroke']\nX_test_0 = np.asarray(X_test_0)\nX_test_0_rank = np.asarray(X_test_0_rank)\ny_test_0 = np.asanyarray(y_test_0)",
   "execution_count": 25,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "model.evaluate(X_test_0, y_test_0)\n",
   "execution_count": 26,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": "200/200 [==============================] - 0s 35us/step\n"
    },
    {
     "data": {
      "text/plain": "[0.0, 1.0]"
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "model_rank.evaluate(X_test_0_rank,y_test_0)\n",
   "execution_count": 27,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": "200/200 [==============================] - 0s 36us/step\n"
    },
    {
     "data": {
      "text/plain": "[0.076906720995903011, 0.90000000000000002]"
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "prediction_0 = model.predict(X_test_0)\nprediction_0_bin = prediction_0 > 0.5\nprediction_0_bin = prediction_0_bin.astype(int)\nconfusion_matrix(y_test_0, prediction_0_bin)",
   "execution_count": 28,
   "outputs": [
    {
     "data": {
      "text/plain": "array([[200]])"
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "prediction_0_rank = model_rank.predict(X_test_0_rank)\nprediction_0_bin_rank = prediction_0_rank > 0.5\nprediction_0_bin_rank = prediction_0_bin_rank.astype(int)\nconfusion_matrix(y_test_0, prediction_0_bin_rank)",
   "execution_count": 29,
   "outputs": [
    {
     "data": {
      "text/plain": "array([[180,  20],\n       [  0,   0]])"
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "plt.figure(figsize=(15,8))\nplt.subplot(121)\nplt.hist(prediction_0)\nplt.title('Histogram prediction')\n\n\nplt.subplot(122)\nplt.hist(prediction_0_rank)\nplt.title('Histogram prediction rank')\nplt.show()",
   "execution_count": 30,
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": "<matplotlib.figure.Figure at 0x1a3091d630>"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "X_test_all = df_test_all.drop(dropColumns,axis=1)\nX_test_all_rank = df_test_all[df_rank.index[1:21]]\ny_test_all = df_test_all['stroke']\nX_test_all = np.asarray(X_test_all)\nX_test_all_rank = np.asarray(X_test_all_rank)\ny_test_all = np.asarray(y_test_all)",
   "execution_count": 31,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "model.evaluate(X_test_all,y_test_all)\n",
   "execution_count": 32,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": "10848/10848 [==============================] - 0s 25us/step\n"
    },
    {
     "data": {
      "text/plain": "[0.015486725663716814, 0.98451327433628322]"
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "model_rank.evaluate(X_test_all_rank, y_test_all)\n",
   "execution_count": 33,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": "10848/10848 [==============================] - 0s 24us/step\n"
    },
    {
     "data": {
      "text/plain": "[0.078372865118591831, 0.89205383480825962]"
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "prediction_all = model.predict(X_test_all)",
   "execution_count": 34,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "prediction_all_rank = model_rank.predict(X_test_all_rank)\n",
   "execution_count": 35,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "plt.figure(figsize=(15,8))\nplt.subplot(121)\nplt.title('Histogram prediction')\nplt.hist(prediction_all)\n\nplt.subplot(122)\nplt.title('Histogram prediction rank')\nplt.hist(prediction_all_rank)\nplt.show()",
   "execution_count": 36,
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": "<matplotlib.figure.Figure at 0x1a309bb8d0>"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "prediction_all_bin = prediction_all > 0.5\nprediction_all_bin = prediction_all_bin.astype(int)\nconfusion_matrix(y_test_all, prediction_all_bin)",
   "execution_count": 37,
   "outputs": [
    {
     "data": {
      "text/plain": "array([[10680,     0],\n       [  168,     0]])"
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "prediction_rank_all_bin = prediction_all_rank > 0.5\nprediction_rank_all_bin = prediction_rank_all_bin.astype(int)\nconfusion_matrix(y_test_all, prediction_rank_all_bin)",
   "execution_count": 38,
   "outputs": [
    {
     "data": {
      "text/plain": "array([[9578, 1102],\n       [  69,   99]])"
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "random_forest = RandomForestClassifier(n_estimators=15,)\nrandom_forest.fit(X_train,y_train)",
   "execution_count": 39,
   "outputs": [
    {
     "data": {
      "text/plain": "RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n            max_depth=None, max_features='auto', max_leaf_nodes=None,\n            min_impurity_decrease=0.0, min_impurity_split=None,\n            min_samples_leaf=1, min_samples_split=2,\n            min_weight_fraction_leaf=0.0, n_estimators=15, n_jobs=1,\n            oob_score=False, random_state=None, verbose=0,\n            warm_start=False)"
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "random_forest.score(X_test,y_test)\n",
   "execution_count": 40,
   "outputs": [
    {
     "data": {
      "text/plain": "0.90833333333333333"
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "prediction_RF = random_forest.predict(X_test)",
   "execution_count": 41,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "confusion_matrix(y_test, prediction_RF)\n",
   "execution_count": 42,
   "outputs": [
    {
     "data": {
      "text/plain": "array([[487,  65],\n       [  1, 167]])"
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "random_forest.score(X_test_0,y_test_0)\n",
   "execution_count": 43,
   "outputs": [
    {
     "data": {
      "text/plain": "0.90000000000000002"
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "prediction_RF_0 = random_forest.predict(X_test_0)\nconfusion_matrix(y_test_0,prediction_RF_0)",
   "execution_count": 44,
   "outputs": [
    {
     "data": {
      "text/plain": "array([[180,  20],\n       [  0,   0]])"
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "random_forest.score(X_test_all,y_test_all)",
   "execution_count": 45,
   "outputs": [
    {
     "data": {
      "text/plain": "0.89657079646017701"
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "prediction_RF_all = random_forest.predict(X_test_all)\nconfusion_matrix(y_test_all, prediction_RF_all)",
   "execution_count": 46,
   "outputs": [
    {
     "data": {
      "text/plain": "array([[9559, 1121],\n       [   1,  167]])"
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "AdaBoost = AdaBoostClassifier(learning_rate=0.1)\nAdaBoost.fit(X_train, y_train)",
   "execution_count": 47,
   "outputs": [
    {
     "data": {
      "text/plain": "AdaBoostClassifier(algorithm='SAMME.R', base_estimator=None,\n          learning_rate=0.1, n_estimators=50, random_state=None)"
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "AdaBoost.score(X_test, y_test)",
   "execution_count": 48,
   "outputs": [
    {
     "data": {
      "text/plain": "0.81944444444444442"
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "prediction_AdaBoost = AdaBoost.predict(X_test)",
   "execution_count": 49,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "confusion_matrix(y_test, prediction_AdaBoost)",
   "execution_count": 50,
   "outputs": [
    {
     "data": {
      "text/plain": "array([[498,  54],\n       [ 76,  92]])"
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "AdaBoost.score(X_test_0,y_test_0)",
   "execution_count": 51,
   "outputs": [
    {
     "data": {
      "text/plain": "0.92500000000000004"
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "prediction_AdaBoost_0 = AdaBoost.predict(X_test_0)\nconfusion_matrix(y_test_0,prediction_AdaBoost_0)\n",
   "execution_count": 52,
   "outputs": [
    {
     "data": {
      "text/plain": "array([[185,  15],\n       [  0,   0]])"
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "AdaBoost.score(X_test_all,y_test_all)",
   "execution_count": 53,
   "outputs": [
    {
     "data": {
      "text/plain": "0.90846238938053092"
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "prediction_AdaBoost_all = AdaBoost.predict(X_test_all)\nconfusion_matrix(y_test_all, prediction_AdaBoost_all)",
   "execution_count": 54,
   "outputs": [
    {
     "data": {
      "text/plain": "array([[9763,  917],\n       [  76,   92]])"
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "classifier = svm.SVC(kernel='linear', C=0.01)\nclassifier.fit(X_train, y_train)\nprediction_SVM = classifier.predict(X_test)",
   "execution_count": 55,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "classifier.score(X_test,y_test)",
   "execution_count": 56,
   "outputs": [
    {
     "data": {
      "text/plain": "0.81805555555555554"
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "confusion_matrix(y_test, prediction_SVM)",
   "execution_count": 57,
   "outputs": [
    {
     "data": {
      "text/plain": "array([[489,  63],\n       [ 68, 100]])"
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "classifier.score(X_test_0,y_test_0)",
   "execution_count": 58,
   "outputs": [
    {
     "data": {
      "text/plain": "0.89000000000000001"
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "prediction_SVM_0 = classifier.predict(X_test_0)\nconfusion_matrix(y_test_0,prediction_SVM_0)",
   "execution_count": 59,
   "outputs": [
    {
     "data": {
      "text/plain": "array([[178,  22],\n       [  0,   0]])"
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "classifier.score(X_test_all,y_test_all)",
   "execution_count": 60,
   "outputs": [
    {
     "data": {
      "text/plain": "0.88938053097345138"
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "prediction_SVM_all = classifier.predict(X_test_all)\nconfusion_matrix(y_test_all, prediction_SVM_all)",
   "execution_count": 61,
   "outputs": [
    {
     "data": {
      "text/plain": "array([[9548, 1132],\n       [  68,  100]])"
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ]
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "",
   "execution_count": null,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "",
   "execution_count": null,
   "outputs": []
  },
  {
   "metadata": {
    "trusted": true
   },
   "cell_type": "code",
   "source": "",
   "execution_count": null,
   "outputs": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "name": "python3",
   "display_name": "Python 3",
   "language": "python"
  },
  "language_info": {
   "name": "python",
   "version": "3.6.4",
   "mimetype": "text/x-python",
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "pygments_lexer": "ipython3",
   "nbconvert_exporter": "python",
   "file_extension": ".py"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}