Final Project - Python Part With Stacking.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "PATH = '/Users/orendar/Downloads/dataverse_files/'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>d.gone.neg.l1</th>\n",
       "      <th>d.gone.neg.l2</th>\n",
       "      <th>d.neg.frac.l3</th>\n",
       "      <th>week</th>\n",
       "      <th>camp.length</th>\n",
       "      <th>base.poll</th>\n",
       "      <th>y2000</th>\n",
       "      <th>y2002</th>\n",
       "      <th>y2004</th>\n",
       "      <th>y2006</th>\n",
       "      <th>...</th>\n",
       "      <th>deminc</th>\n",
       "      <th>dem.contrib.l1</th>\n",
       "      <th>dem.contrib.l2</th>\n",
       "      <th>rep.contrib.l1</th>\n",
       "      <th>rep.contrib.l2</th>\n",
       "      <th>d.gone.neg</th>\n",
       "      <th>gbc_preds</th>\n",
       "      <th>rfc_preds</th>\n",
       "      <th>etc_preds</th>\n",
       "      <th>nn_preds</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-11</td>\n",
       "      <td>23</td>\n",
       "      <td>45.744681</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>38350.0</td>\n",
       "      <td>126536.00</td>\n",
       "      <td>132545.00</td>\n",
       "      <td>102559.36</td>\n",
       "      <td>0</td>\n",
       "      <td>0.235871</td>\n",
       "      <td>0.146</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.199028</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-10</td>\n",
       "      <td>23</td>\n",
       "      <td>45.744681</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>58035.0</td>\n",
       "      <td>38350.00</td>\n",
       "      <td>522327.00</td>\n",
       "      <td>132545.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0.154327</td>\n",
       "      <td>0.084</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.283316</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-8</td>\n",
       "      <td>23</td>\n",
       "      <td>45.744681</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>167280.0</td>\n",
       "      <td>67068.67</td>\n",
       "      <td>177739.44</td>\n",
       "      <td>138533.00</td>\n",
       "      <td>1</td>\n",
       "      <td>0.921011</td>\n",
       "      <td>0.958</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.906001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-7</td>\n",
       "      <td>23</td>\n",
       "      <td>45.744681</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>389947.0</td>\n",
       "      <td>167280.00</td>\n",
       "      <td>543465.00</td>\n",
       "      <td>177739.44</td>\n",
       "      <td>1</td>\n",
       "      <td>0.985290</td>\n",
       "      <td>0.998</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.974159</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.071429</td>\n",
       "      <td>-6</td>\n",
       "      <td>23</td>\n",
       "      <td>45.744681</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>202734.0</td>\n",
       "      <td>389947.00</td>\n",
       "      <td>2415314.50</td>\n",
       "      <td>543465.00</td>\n",
       "      <td>1</td>\n",
       "      <td>0.989649</td>\n",
       "      <td>1.000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.994450</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 34 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   d.gone.neg.l1  d.gone.neg.l2  d.neg.frac.l3  week  camp.length  base.poll  \\\n",
       "1              0              0       0.000000   -11           23  45.744681   \n",
       "2              0              0       0.000000   -10           23  45.744681   \n",
       "3              1              0       0.000000    -8           23  45.744681   \n",
       "4              1              1       0.000000    -7           23  45.744681   \n",
       "5              1              1       0.071429    -6           23  45.744681   \n",
       "\n",
       "   y2000  y2002  y2004  y2006    ...     deminc  dem.contrib.l1  \\\n",
       "1      0      1      0      0    ...          1         38350.0   \n",
       "2      0      1      0      0    ...          1         58035.0   \n",
       "3      0      1      0      0    ...          1        167280.0   \n",
       "4      0      1      0      0    ...          1        389947.0   \n",
       "5      0      1      0      0    ...          1        202734.0   \n",
       "\n",
       "   dem.contrib.l2  rep.contrib.l1  rep.contrib.l2  d.gone.neg  gbc_preds  \\\n",
       "1       126536.00       132545.00       102559.36           0   0.235871   \n",
       "2        38350.00       522327.00       132545.00           0   0.154327   \n",
       "3        67068.67       177739.44       138533.00           1   0.921011   \n",
       "4       167280.00       543465.00       177739.44           1   0.985290   \n",
       "5       389947.00      2415314.50       543465.00           1   0.989649   \n",
       "\n",
       "   rfc_preds  etc_preds  nn_preds  \n",
       "1      0.146        0.0  0.199028  \n",
       "2      0.084        0.0  0.283316  \n",
       "3      0.958        1.0  0.906001  \n",
       "4      0.998        1.0  0.974159  \n",
       "5      1.000        1.0  0.994450  \n",
       "\n",
       "[5 rows x 34 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df = pd.read_csv(f'{PATH}train_preds.csv', index_col = 0)\n",
    "train_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>d.gone.neg.l1</th>\n",
       "      <th>d.gone.neg.l2</th>\n",
       "      <th>d.neg.frac.l3</th>\n",
       "      <th>week</th>\n",
       "      <th>camp.length</th>\n",
       "      <th>base.poll</th>\n",
       "      <th>y2000</th>\n",
       "      <th>y2002</th>\n",
       "      <th>y2004</th>\n",
       "      <th>y2006</th>\n",
       "      <th>...</th>\n",
       "      <th>deminc</th>\n",
       "      <th>dem.contrib.l1</th>\n",
       "      <th>dem.contrib.l2</th>\n",
       "      <th>rep.contrib.l1</th>\n",
       "      <th>rep.contrib.l2</th>\n",
       "      <th>d.gone.neg</th>\n",
       "      <th>gbc_preds</th>\n",
       "      <th>rfc_preds</th>\n",
       "      <th>etc_preds</th>\n",
       "      <th>nn_preds</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-7</td>\n",
       "      <td>13</td>\n",
       "      <td>44.736842</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>41083.00</td>\n",
       "      <td>61295.00</td>\n",
       "      <td>289158.00</td>\n",
       "      <td>344029.00</td>\n",
       "      <td>1</td>\n",
       "      <td>0.944583</td>\n",
       "      <td>0.882</td>\n",
       "      <td>0.942</td>\n",
       "      <td>0.900369</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-6</td>\n",
       "      <td>23</td>\n",
       "      <td>63.636364</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>699168.06</td>\n",
       "      <td>175919.18</td>\n",
       "      <td>34550.00</td>\n",
       "      <td>8641.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0.020580</td>\n",
       "      <td>0.006</td>\n",
       "      <td>0.010</td>\n",
       "      <td>0.020276</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.176471</td>\n",
       "      <td>-3</td>\n",
       "      <td>23</td>\n",
       "      <td>62.765957</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>383312.88</td>\n",
       "      <td>420975.89</td>\n",
       "      <td>31510.13</td>\n",
       "      <td>390706.48</td>\n",
       "      <td>1</td>\n",
       "      <td>0.902891</td>\n",
       "      <td>0.920</td>\n",
       "      <td>0.962</td>\n",
       "      <td>0.946103</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.090909</td>\n",
       "      <td>-2</td>\n",
       "      <td>27</td>\n",
       "      <td>62.637363</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>57150.00</td>\n",
       "      <td>34750.00</td>\n",
       "      <td>219018.00</td>\n",
       "      <td>194750.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0.029230</td>\n",
       "      <td>0.156</td>\n",
       "      <td>0.078</td>\n",
       "      <td>0.028789</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-6</td>\n",
       "      <td>15</td>\n",
       "      <td>66.455237</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>165910.00</td>\n",
       "      <td>175460.08</td>\n",
       "      <td>91269.03</td>\n",
       "      <td>66217.39</td>\n",
       "      <td>0</td>\n",
       "      <td>0.151891</td>\n",
       "      <td>0.246</td>\n",
       "      <td>0.180</td>\n",
       "      <td>0.121091</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 34 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   d.gone.neg.l1  d.gone.neg.l2  d.neg.frac.l3  week  camp.length  base.poll  \\\n",
       "1              1              1       1.000000    -7           13  44.736842   \n",
       "2              0              0       0.000000    -6           23  63.636364   \n",
       "3              1              1       0.176471    -3           23  62.765957   \n",
       "4              0              0       0.090909    -2           27  62.637363   \n",
       "5              0              0       0.000000    -6           15  66.455237   \n",
       "\n",
       "   y2000  y2002  y2004  y2006    ...     deminc  dem.contrib.l1  \\\n",
       "1      0      1      0      0    ...          0        41083.00   \n",
       "2      0      0      0      1    ...          1       699168.06   \n",
       "3      0      1      0      0    ...          0       383312.88   \n",
       "4      0      0      0      1    ...          1        57150.00   \n",
       "5      0      0      0      1    ...          1       165910.00   \n",
       "\n",
       "   dem.contrib.l2  rep.contrib.l1  rep.contrib.l2  d.gone.neg  gbc_preds  \\\n",
       "1        61295.00       289158.00       344029.00           1   0.944583   \n",
       "2       175919.18        34550.00         8641.00           0   0.020580   \n",
       "3       420975.89        31510.13       390706.48           1   0.902891   \n",
       "4        34750.00       219018.00       194750.00           0   0.029230   \n",
       "5       175460.08        91269.03        66217.39           0   0.151891   \n",
       "\n",
       "   rfc_preds  etc_preds  nn_preds  \n",
       "1      0.882      0.942  0.900369  \n",
       "2      0.006      0.010  0.020276  \n",
       "3      0.920      0.962  0.946103  \n",
       "4      0.156      0.078  0.028789  \n",
       "5      0.246      0.180  0.121091  \n",
       "\n",
       "[5 rows x 34 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_df = pd.read_csv(f'{PATH}test_preds.csv',  index_col = 0)\n",
    "test_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((818, 33), (818,), (274, 33), (274,))"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "target = \"d.gone.neg\"\n",
    "\n",
    "y_train = train_df[target]\n",
    "x_train = train_df.drop(target, axis=1)\n",
    "y_test = test_df[target]\n",
    "x_test = test_df.drop(target, axis=1)\n",
    "x_train.shape, y_train.shape, x_test.shape, y_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import *\n",
    "from sklearn.metrics import *\n",
    "rs = 42"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "gbc = GradientBoostingClassifier(n_estimators=100, max_features='sqrt', random_state=rs)\n",
    "rfc = RandomForestClassifier(n_estimators=500, random_state=rs, n_jobs=-1)\n",
    "etc = ExtraTreesClassifier(n_estimators=500, random_state=rs, n_jobs=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "gbc.fit(x_train, y_train)\n",
    "rfc.fit(x_train, y_train)\n",
    "etc.fit(x_train, y_train);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "gbc_preds = gbc.predict(x_test)\n",
    "rfc_preds = rfc.predict(x_test)\n",
    "etc_preds = etc.predict(x_test)\n",
    "gbc_prob_preds = gbc.predict_proba(x_test)\n",
    "rfc_prob_preds = rfc.predict_proba(x_test)\n",
    "etc_prob_preds = etc.predict_proba(x_test)\n",
    "gbc_pos = gbc_prob_preds.T[1]\n",
    "rfc_pos = rfc_prob_preds.T[1]\n",
    "etc_pos = etc_prob_preds.T[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def rmse(preds): return mean_squared_error(y_test, preds)**.5\n",
    "def auroc(preds): return roc_auc_score(y_test, preds)\n",
    "def accuracy(preds): return accuracy_score(y_test, preds)\n",
    "def logloss(preds): return log_loss(y_test, preds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.3601286977826547, 0.34268785140948554, 0.3012700851952889)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rmse(gbc_pos), rmse(rfc_pos), rmse(etc_pos)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.864963503649635, 0.864963503649635, 0.8576642335766423)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy(gbc_preds), accuracy(rfc_preds), accuracy(etc_preds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.9356060606060606, 0.9272360703812317, 0.9376221896383187)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "auroc(gbc_pos), auroc(rfc_pos), auroc(etc_pos)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1.032127670694763, 0.7937493292502644, 0.29190121457875523)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "logloss(gbc_pos), logloss(rfc_pos), logloss(etc_pos)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "scrolled": true
   },
   "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>Accuracy</th>\n",
       "      <th>RMSE</th>\n",
       "      <th>AUROC</th>\n",
       "      <th>LogLoss</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>GBM</th>\n",
       "      <td>0.864964</td>\n",
       "      <td>0.360129</td>\n",
       "      <td>0.935606</td>\n",
       "      <td>1.032128</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RF</th>\n",
       "      <td>0.864964</td>\n",
       "      <td>0.342688</td>\n",
       "      <td>0.927236</td>\n",
       "      <td>0.793749</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ET</th>\n",
       "      <td>0.857664</td>\n",
       "      <td>0.301270</td>\n",
       "      <td>0.937622</td>\n",
       "      <td>0.291901</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Accuracy      RMSE     AUROC   LogLoss\n",
       "GBM  0.864964  0.360129  0.935606  1.032128\n",
       "RF   0.864964  0.342688  0.927236  0.793749\n",
       "ET   0.857664  0.301270  0.937622  0.291901"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame(data=[[accuracy(gbc_preds), rmse(gbc_pos), auroc(gbc_pos), logloss(gbc_pos)],\n",
    "                        [accuracy(rfc_preds), rmse(rfc_pos), auroc(rfc_pos), logloss(rfc_pos)],\n",
    "                        [accuracy(etc_preds), rmse(etc_pos), auroc(etc_pos), logloss(etc_pos)]],\n",
    "                  index=['GBM', 'RF', 'ET'],\n",
    "                  columns=['Accuracy', 'RMSE', 'AUROC', 'LogLoss'])\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'\\\\begin{tabular}{lrrrr}\\n\\\\toprule\\n{} &  Accuracy &      RMSE &     AUROC &   LogLoss \\\\\\\\\\n\\\\midrule\\nGBM &  0.864964 &  0.360129 &  0.935606 &  1.032128 \\\\\\\\\\nRF  &  0.864964 &  0.342688 &  0.927236 &  0.793749 \\\\\\\\\\nET  &  0.857664 &  0.301270 &  0.937622 &  0.291901 \\\\\\\\\\n\\\\bottomrule\\n\\\\end{tabular}\\n'"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.to_latex()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/orendar/anaconda3/lib/python3.6/site-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n",
      "  from ._conv import register_converters as _register_converters\n",
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from keras.models import *\n",
    "from keras.layers import *\n",
    "from keras.utils import *\n",
    "from keras.optimizers import *\n",
    "from clr import *\n",
    "from sklearn.preprocessing import *\n",
    "np.random.seed(42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train_mat, y_train_mat = x_train.as_matrix(), y_train.as_matrix().reshape(-1, 1)\n",
    "x_test_mat, y_test_mat = x_test.as_matrix(), y_test.as_matrix().reshape(-1, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = StandardScaler()\n",
    "x_train_norm = scaler.fit_transform(x_train_mat)\n",
    "x_test_norm = scaler.transform(x_test_mat)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "model.add(Dropout(0.15, input_shape=(33,)))\n",
    "model.add(Dense(4096, activation='relu'))\n",
    "model.add(Dropout(0.5))\n",
    "model.add(Dense(1, activation='sigmoid'))\n",
    "\n",
    "# Compile model\n",
    "opt = optimizers.adam(lr=5e-6, beta_1=0.9, beta_2=0.999, epsilon=None, decay=0.0, amsgrad=True)\n",
    "#opt = optimizers.SGD(lr=5e-4, momentum=0.9, decay=0, nesterov=True)\n",
    "model.compile(loss='binary_crossentropy', optimizer=opt, metrics=['mse', 'accuracy'])\n",
    "\n",
    "#callbacks arsenal\n",
    "def sched(idx, lr): return lr*0.9 if idx % 10 == 0 else lr\n",
    "lrsched = LearningRateScheduler(sched, verbose=0)\n",
    "reducelr = ReduceLROnPlateau(factor=0.8, patience=2)\n",
    "stop = EarlyStopping(patience=50)\n",
    "clr = CyclicLR(base_lr=1e-6, max_lr=1e-5, step_size=20, mode='triangular', scale_mode='cycle')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 818 samples, validate on 274 samples\n",
      "Epoch 1/3000\n",
      " - 0s - loss: 0.6926 - mean_squared_error: 0.2498 - acc: 0.5098 - val_loss: 0.6958 - val_mean_squared_error: 0.2513 - val_acc: 0.4489\n",
      "Epoch 2/3000\n",
      " - 0s - loss: 0.7007 - mean_squared_error: 0.2538 - acc: 0.4609 - val_loss: 0.6948 - val_mean_squared_error: 0.2508 - val_acc: 0.4599\n",
      "Epoch 3/3000\n",
      " - 0s - loss: 0.6975 - mean_squared_error: 0.2521 - acc: 0.4927 - val_loss: 0.6935 - val_mean_squared_error: 0.2502 - val_acc: 0.4818\n",
      "Epoch 4/3000\n",
      " - 0s - loss: 0.6928 - mean_squared_error: 0.2498 - acc: 0.5061 - val_loss: 0.6918 - val_mean_squared_error: 0.2493 - val_acc: 0.4891\n",
      "Epoch 5/3000\n",
      " - 0s - loss: 0.6941 - mean_squared_error: 0.2505 - acc: 0.4988 - val_loss: 0.6896 - val_mean_squared_error: 0.2483 - val_acc: 0.4964\n",
      "Epoch 6/3000\n",
      " - 0s - loss: 0.6896 - mean_squared_error: 0.2482 - acc: 0.5269 - val_loss: 0.6872 - val_mean_squared_error: 0.2470 - val_acc: 0.5073\n",
      "Epoch 7/3000\n",
      " - 0s - loss: 0.6843 - mean_squared_error: 0.2456 - acc: 0.5489 - val_loss: 0.6843 - val_mean_squared_error: 0.2456 - val_acc: 0.5328\n",
      "Epoch 8/3000\n",
      " - 0s - loss: 0.6799 - mean_squared_error: 0.2435 - acc: 0.5550 - val_loss: 0.6811 - val_mean_squared_error: 0.2440 - val_acc: 0.5584\n",
      "Epoch 9/3000\n",
      " - 0s - loss: 0.6806 - mean_squared_error: 0.2438 - acc: 0.5599 - val_loss: 0.6775 - val_mean_squared_error: 0.2422 - val_acc: 0.5949\n",
      "Epoch 10/3000\n",
      " - 0s - loss: 0.6809 - mean_squared_error: 0.2440 - acc: 0.5538 - val_loss: 0.6736 - val_mean_squared_error: 0.2403 - val_acc: 0.6168\n",
      "Epoch 11/3000\n",
      " - 0s - loss: 0.6720 - mean_squared_error: 0.2395 - acc: 0.6137 - val_loss: 0.6696 - val_mean_squared_error: 0.2383 - val_acc: 0.6423\n",
      "Epoch 12/3000\n",
      " - 0s - loss: 0.6662 - mean_squared_error: 0.2367 - acc: 0.6308 - val_loss: 0.6660 - val_mean_squared_error: 0.2365 - val_acc: 0.6715\n",
      "Epoch 13/3000\n",
      " - 0s - loss: 0.6595 - mean_squared_error: 0.2334 - acc: 0.6540 - val_loss: 0.6628 - val_mean_squared_error: 0.2349 - val_acc: 0.6934\n",
      "Epoch 14/3000\n",
      " - 0s - loss: 0.6565 - mean_squared_error: 0.2319 - acc: 0.6785 - val_loss: 0.6600 - val_mean_squared_error: 0.2335 - val_acc: 0.7190\n",
      "Epoch 15/3000\n",
      " - 0s - loss: 0.6536 - mean_squared_error: 0.2305 - acc: 0.6601 - val_loss: 0.6575 - val_mean_squared_error: 0.2323 - val_acc: 0.7263\n",
      "Epoch 16/3000\n",
      " - 0s - loss: 0.6556 - mean_squared_error: 0.2314 - acc: 0.7017 - val_loss: 0.6554 - val_mean_squared_error: 0.2313 - val_acc: 0.7372\n",
      "Epoch 17/3000\n",
      " - 0s - loss: 0.6475 - mean_squared_error: 0.2274 - acc: 0.7042 - val_loss: 0.6537 - val_mean_squared_error: 0.2304 - val_acc: 0.7336\n",
      "Epoch 18/3000\n",
      " - 0s - loss: 0.6523 - mean_squared_error: 0.2299 - acc: 0.6797 - val_loss: 0.6524 - val_mean_squared_error: 0.2298 - val_acc: 0.7372\n",
      "Epoch 19/3000\n",
      " - 0s - loss: 0.6541 - mean_squared_error: 0.2307 - acc: 0.6675 - val_loss: 0.6514 - val_mean_squared_error: 0.2293 - val_acc: 0.7445\n",
      "Epoch 20/3000\n",
      " - 0s - loss: 0.6468 - mean_squared_error: 0.2271 - acc: 0.7042 - val_loss: 0.6507 - val_mean_squared_error: 0.2289 - val_acc: 0.7482\n",
      "Epoch 21/3000\n",
      " - 0s - loss: 0.6471 - mean_squared_error: 0.2273 - acc: 0.6944 - val_loss: 0.6503 - val_mean_squared_error: 0.2287 - val_acc: 0.7518\n",
      "Epoch 22/3000\n",
      " - 0s - loss: 0.6475 - mean_squared_error: 0.2274 - acc: 0.7066 - val_loss: 0.6495 - val_mean_squared_error: 0.2283 - val_acc: 0.7555\n",
      "Epoch 23/3000\n",
      " - 0s - loss: 0.6426 - mean_squared_error: 0.2250 - acc: 0.7359 - val_loss: 0.6483 - val_mean_squared_error: 0.2277 - val_acc: 0.7591\n",
      "Epoch 24/3000\n",
      " - 0s - loss: 0.6395 - mean_squared_error: 0.2235 - acc: 0.7262 - val_loss: 0.6468 - val_mean_squared_error: 0.2270 - val_acc: 0.7628\n",
      "Epoch 25/3000\n",
      " - 0s - loss: 0.6389 - mean_squared_error: 0.2233 - acc: 0.7274 - val_loss: 0.6450 - val_mean_squared_error: 0.2261 - val_acc: 0.7737\n",
      "Epoch 26/3000\n",
      " - 0s - loss: 0.6385 - mean_squared_error: 0.2230 - acc: 0.7335 - val_loss: 0.6429 - val_mean_squared_error: 0.2251 - val_acc: 0.7737\n",
      "Epoch 27/3000\n",
      " - 0s - loss: 0.6354 - mean_squared_error: 0.2215 - acc: 0.7396 - val_loss: 0.6404 - val_mean_squared_error: 0.2239 - val_acc: 0.7883\n",
      "Epoch 28/3000\n",
      " - 0s - loss: 0.6348 - mean_squared_error: 0.2212 - acc: 0.7396 - val_loss: 0.6376 - val_mean_squared_error: 0.2225 - val_acc: 0.7847\n",
      "Epoch 29/3000\n",
      " - 0s - loss: 0.6308 - mean_squared_error: 0.2192 - acc: 0.7555 - val_loss: 0.6346 - val_mean_squared_error: 0.2210 - val_acc: 0.7883\n",
      "Epoch 30/3000\n",
      " - 0s - loss: 0.6297 - mean_squared_error: 0.2187 - acc: 0.7494 - val_loss: 0.6312 - val_mean_squared_error: 0.2193 - val_acc: 0.7956\n",
      "Epoch 31/3000\n",
      " - 0s - loss: 0.6245 - mean_squared_error: 0.2161 - acc: 0.7702 - val_loss: 0.6277 - val_mean_squared_error: 0.2176 - val_acc: 0.8029\n",
      "Epoch 32/3000\n",
      " - 0s - loss: 0.6174 - mean_squared_error: 0.2127 - acc: 0.7775 - val_loss: 0.6245 - val_mean_squared_error: 0.2161 - val_acc: 0.8066\n",
      "Epoch 33/3000\n",
      " - 0s - loss: 0.6146 - mean_squared_error: 0.2113 - acc: 0.7983 - val_loss: 0.6218 - val_mean_squared_error: 0.2147 - val_acc: 0.8066\n",
      "Epoch 34/3000\n",
      " - 0s - loss: 0.6140 - mean_squared_error: 0.2110 - acc: 0.7995 - val_loss: 0.6193 - val_mean_squared_error: 0.2136 - val_acc: 0.8102\n",
      "Epoch 35/3000\n",
      " - 0s - loss: 0.6098 - mean_squared_error: 0.2090 - acc: 0.8056 - val_loss: 0.6173 - val_mean_squared_error: 0.2125 - val_acc: 0.8066\n",
      "Epoch 36/3000\n",
      " - 0s - loss: 0.6077 - mean_squared_error: 0.2080 - acc: 0.8068 - val_loss: 0.6155 - val_mean_squared_error: 0.2117 - val_acc: 0.8102\n",
      "Epoch 37/3000\n",
      " - 0s - loss: 0.6078 - mean_squared_error: 0.2080 - acc: 0.8044 - val_loss: 0.6140 - val_mean_squared_error: 0.2110 - val_acc: 0.8102\n",
      "Epoch 38/3000\n",
      " - 0s - loss: 0.6043 - mean_squared_error: 0.2064 - acc: 0.8227 - val_loss: 0.6129 - val_mean_squared_error: 0.2104 - val_acc: 0.8102\n",
      "Epoch 39/3000\n",
      " - 0s - loss: 0.6030 - mean_squared_error: 0.2058 - acc: 0.8191 - val_loss: 0.6121 - val_mean_squared_error: 0.2100 - val_acc: 0.8139\n",
      "Epoch 40/3000\n",
      " - 0s - loss: 0.6048 - mean_squared_error: 0.2066 - acc: 0.8154 - val_loss: 0.6115 - val_mean_squared_error: 0.2098 - val_acc: 0.8212\n",
      "Epoch 41/3000\n",
      " - 0s - loss: 0.6030 - mean_squared_error: 0.2057 - acc: 0.8081 - val_loss: 0.6111 - val_mean_squared_error: 0.2096 - val_acc: 0.8212\n",
      "Epoch 42/3000\n",
      " - 0s - loss: 0.6018 - mean_squared_error: 0.2052 - acc: 0.8044 - val_loss: 0.6104 - val_mean_squared_error: 0.2092 - val_acc: 0.8212\n",
      "Epoch 43/3000\n",
      " - 0s - loss: 0.6043 - mean_squared_error: 0.2063 - acc: 0.8105 - val_loss: 0.6094 - val_mean_squared_error: 0.2087 - val_acc: 0.8248\n",
      "Epoch 44/3000\n",
      " - 0s - loss: 0.5990 - mean_squared_error: 0.2038 - acc: 0.8178 - val_loss: 0.6082 - val_mean_squared_error: 0.2081 - val_acc: 0.8248\n",
      "Epoch 45/3000\n",
      " - 0s - loss: 0.6016 - mean_squared_error: 0.2050 - acc: 0.7983 - val_loss: 0.6066 - val_mean_squared_error: 0.2074 - val_acc: 0.8248\n",
      "Epoch 46/3000\n",
      " - 0s - loss: 0.5989 - mean_squared_error: 0.2038 - acc: 0.8227 - val_loss: 0.6048 - val_mean_squared_error: 0.2065 - val_acc: 0.8248\n",
      "Epoch 47/3000\n",
      " - 0s - loss: 0.5924 - mean_squared_error: 0.2006 - acc: 0.8068 - val_loss: 0.6027 - val_mean_squared_error: 0.2055 - val_acc: 0.8285\n",
      "Epoch 48/3000\n",
      " - 0s - loss: 0.5921 - mean_squared_error: 0.2005 - acc: 0.8203 - val_loss: 0.6003 - val_mean_squared_error: 0.2044 - val_acc: 0.8285\n",
      "Epoch 49/3000\n",
      " - 0s - loss: 0.5911 - mean_squared_error: 0.1998 - acc: 0.8386 - val_loss: 0.5977 - val_mean_squared_error: 0.2031 - val_acc: 0.8285\n",
      "Epoch 50/3000\n",
      " - 0s - loss: 0.5858 - mean_squared_error: 0.1974 - acc: 0.8325 - val_loss: 0.5948 - val_mean_squared_error: 0.2017 - val_acc: 0.8321\n",
      "Epoch 51/3000\n",
      " - 0s - loss: 0.5834 - mean_squared_error: 0.1963 - acc: 0.8399 - val_loss: 0.5918 - val_mean_squared_error: 0.2003 - val_acc: 0.8358\n",
      "Epoch 52/3000\n",
      " - 0s - loss: 0.5776 - mean_squared_error: 0.1936 - acc: 0.8411 - val_loss: 0.5892 - val_mean_squared_error: 0.1990 - val_acc: 0.8394\n",
      "Epoch 53/3000\n",
      " - 0s - loss: 0.5779 - mean_squared_error: 0.1937 - acc: 0.8276 - val_loss: 0.5868 - val_mean_squared_error: 0.1979 - val_acc: 0.8431\n",
      "Epoch 54/3000\n",
      " - 0s - loss: 0.5761 - mean_squared_error: 0.1927 - acc: 0.8582 - val_loss: 0.5847 - val_mean_squared_error: 0.1969 - val_acc: 0.8467\n",
      "Epoch 55/3000\n",
      " - 0s - loss: 0.5746 - mean_squared_error: 0.1921 - acc: 0.8435 - val_loss: 0.5829 - val_mean_squared_error: 0.1961 - val_acc: 0.8467\n",
      "Epoch 56/3000\n",
      " - 0s - loss: 0.5709 - mean_squared_error: 0.1904 - acc: 0.8423 - val_loss: 0.5814 - val_mean_squared_error: 0.1954 - val_acc: 0.8504\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 57/3000\n",
      " - 0s - loss: 0.5700 - mean_squared_error: 0.1900 - acc: 0.8325 - val_loss: 0.5801 - val_mean_squared_error: 0.1948 - val_acc: 0.8504\n",
      "Epoch 58/3000\n",
      " - 0s - loss: 0.5667 - mean_squared_error: 0.1883 - acc: 0.8509 - val_loss: 0.5791 - val_mean_squared_error: 0.1943 - val_acc: 0.8540\n",
      "Epoch 59/3000\n",
      " - 0s - loss: 0.5645 - mean_squared_error: 0.1876 - acc: 0.8337 - val_loss: 0.5784 - val_mean_squared_error: 0.1940 - val_acc: 0.8540\n",
      "Epoch 60/3000\n",
      " - 0s - loss: 0.5687 - mean_squared_error: 0.1894 - acc: 0.8545 - val_loss: 0.5779 - val_mean_squared_error: 0.1937 - val_acc: 0.8540\n",
      "Epoch 61/3000\n",
      " - 0s - loss: 0.5640 - mean_squared_error: 0.1871 - acc: 0.8716 - val_loss: 0.5776 - val_mean_squared_error: 0.1936 - val_acc: 0.8540\n",
      "Epoch 62/3000\n",
      " - 0s - loss: 0.5685 - mean_squared_error: 0.1893 - acc: 0.8533 - val_loss: 0.5770 - val_mean_squared_error: 0.1933 - val_acc: 0.8540\n",
      "Epoch 63/3000\n",
      " - 0s - loss: 0.5589 - mean_squared_error: 0.1847 - acc: 0.8643 - val_loss: 0.5762 - val_mean_squared_error: 0.1929 - val_acc: 0.8540\n",
      "Epoch 64/3000\n",
      " - 0s - loss: 0.5638 - mean_squared_error: 0.1870 - acc: 0.8533 - val_loss: 0.5751 - val_mean_squared_error: 0.1924 - val_acc: 0.8540\n",
      "Epoch 65/3000\n",
      " - 0s - loss: 0.5647 - mean_squared_error: 0.1874 - acc: 0.8313 - val_loss: 0.5737 - val_mean_squared_error: 0.1918 - val_acc: 0.8540\n",
      "Epoch 66/3000\n",
      " - 0s - loss: 0.5603 - mean_squared_error: 0.1852 - acc: 0.8643 - val_loss: 0.5722 - val_mean_squared_error: 0.1910 - val_acc: 0.8540\n",
      "Epoch 67/3000\n",
      " - 0s - loss: 0.5555 - mean_squared_error: 0.1831 - acc: 0.8680 - val_loss: 0.5704 - val_mean_squared_error: 0.1902 - val_acc: 0.8540\n",
      "Epoch 68/3000\n",
      " - 0s - loss: 0.5573 - mean_squared_error: 0.1840 - acc: 0.8521 - val_loss: 0.5684 - val_mean_squared_error: 0.1892 - val_acc: 0.8540\n",
      "Epoch 69/3000\n",
      " - 0s - loss: 0.5538 - mean_squared_error: 0.1823 - acc: 0.8582 - val_loss: 0.5661 - val_mean_squared_error: 0.1882 - val_acc: 0.8540\n",
      "Epoch 70/3000\n",
      " - 0s - loss: 0.5543 - mean_squared_error: 0.1825 - acc: 0.8533 - val_loss: 0.5637 - val_mean_squared_error: 0.1871 - val_acc: 0.8577\n",
      "Epoch 71/3000\n",
      " - 0s - loss: 0.5461 - mean_squared_error: 0.1787 - acc: 0.8667 - val_loss: 0.5611 - val_mean_squared_error: 0.1859 - val_acc: 0.8577\n",
      "Epoch 72/3000\n",
      " - 0s - loss: 0.5440 - mean_squared_error: 0.1778 - acc: 0.8509 - val_loss: 0.5589 - val_mean_squared_error: 0.1848 - val_acc: 0.8577\n",
      "Epoch 73/3000\n",
      " - 0s - loss: 0.5405 - mean_squared_error: 0.1760 - acc: 0.8667 - val_loss: 0.5568 - val_mean_squared_error: 0.1839 - val_acc: 0.8577\n",
      "Epoch 74/3000\n",
      " - 0s - loss: 0.5389 - mean_squared_error: 0.1753 - acc: 0.8655 - val_loss: 0.5551 - val_mean_squared_error: 0.1831 - val_acc: 0.8577\n",
      "Epoch 75/3000\n",
      " - 0s - loss: 0.5361 - mean_squared_error: 0.1741 - acc: 0.8631 - val_loss: 0.5535 - val_mean_squared_error: 0.1824 - val_acc: 0.8577\n",
      "Epoch 76/3000\n",
      " - 0s - loss: 0.5342 - mean_squared_error: 0.1731 - acc: 0.8667 - val_loss: 0.5522 - val_mean_squared_error: 0.1818 - val_acc: 0.8577\n",
      "Epoch 77/3000\n",
      " - 0s - loss: 0.5346 - mean_squared_error: 0.1734 - acc: 0.8826 - val_loss: 0.5511 - val_mean_squared_error: 0.1813 - val_acc: 0.8577\n",
      "Epoch 78/3000\n",
      " - 0s - loss: 0.5314 - mean_squared_error: 0.1719 - acc: 0.8924 - val_loss: 0.5503 - val_mean_squared_error: 0.1809 - val_acc: 0.8577\n",
      "Epoch 79/3000\n",
      " - 0s - loss: 0.5330 - mean_squared_error: 0.1727 - acc: 0.8643 - val_loss: 0.5496 - val_mean_squared_error: 0.1806 - val_acc: 0.8577\n",
      "Epoch 80/3000\n",
      " - 0s - loss: 0.5292 - mean_squared_error: 0.1709 - acc: 0.8716 - val_loss: 0.5492 - val_mean_squared_error: 0.1804 - val_acc: 0.8577\n",
      "Epoch 81/3000\n",
      " - 0s - loss: 0.5374 - mean_squared_error: 0.1749 - acc: 0.8667 - val_loss: 0.5489 - val_mean_squared_error: 0.1803 - val_acc: 0.8577\n",
      "Epoch 82/3000\n",
      " - 0s - loss: 0.5346 - mean_squared_error: 0.1736 - acc: 0.8496 - val_loss: 0.5484 - val_mean_squared_error: 0.1801 - val_acc: 0.8577\n",
      "Epoch 83/3000\n",
      " - 0s - loss: 0.5293 - mean_squared_error: 0.1709 - acc: 0.8753 - val_loss: 0.5477 - val_mean_squared_error: 0.1797 - val_acc: 0.8577\n",
      "Epoch 84/3000\n",
      " - 0s - loss: 0.5303 - mean_squared_error: 0.1715 - acc: 0.8643 - val_loss: 0.5468 - val_mean_squared_error: 0.1793 - val_acc: 0.8577\n",
      "Epoch 85/3000\n",
      " - 0s - loss: 0.5275 - mean_squared_error: 0.1702 - acc: 0.8729 - val_loss: 0.5456 - val_mean_squared_error: 0.1788 - val_acc: 0.8577\n",
      "Epoch 86/3000\n",
      " - 0s - loss: 0.5272 - mean_squared_error: 0.1700 - acc: 0.8643 - val_loss: 0.5443 - val_mean_squared_error: 0.1782 - val_acc: 0.8613\n",
      "Epoch 87/3000\n",
      " - 0s - loss: 0.5274 - mean_squared_error: 0.1701 - acc: 0.8643 - val_loss: 0.5427 - val_mean_squared_error: 0.1775 - val_acc: 0.8613\n",
      "Epoch 88/3000\n",
      " - 0s - loss: 0.5244 - mean_squared_error: 0.1688 - acc: 0.8729 - val_loss: 0.5410 - val_mean_squared_error: 0.1767 - val_acc: 0.8613\n",
      "Epoch 89/3000\n",
      " - 0s - loss: 0.5228 - mean_squared_error: 0.1681 - acc: 0.8729 - val_loss: 0.5390 - val_mean_squared_error: 0.1758 - val_acc: 0.8613\n",
      "Epoch 90/3000\n",
      " - 0s - loss: 0.5204 - mean_squared_error: 0.1671 - acc: 0.8680 - val_loss: 0.5369 - val_mean_squared_error: 0.1748 - val_acc: 0.8577\n",
      "Epoch 91/3000\n",
      " - 0s - loss: 0.5139 - mean_squared_error: 0.1638 - acc: 0.8814 - val_loss: 0.5347 - val_mean_squared_error: 0.1738 - val_acc: 0.8613\n",
      "Epoch 92/3000\n",
      " - 0s - loss: 0.5122 - mean_squared_error: 0.1632 - acc: 0.8802 - val_loss: 0.5327 - val_mean_squared_error: 0.1729 - val_acc: 0.8613\n",
      "Epoch 93/3000\n",
      " - 0s - loss: 0.5117 - mean_squared_error: 0.1630 - acc: 0.8826 - val_loss: 0.5309 - val_mean_squared_error: 0.1722 - val_acc: 0.8613\n",
      "Epoch 94/3000\n",
      " - 0s - loss: 0.5112 - mean_squared_error: 0.1629 - acc: 0.8790 - val_loss: 0.5294 - val_mean_squared_error: 0.1715 - val_acc: 0.8577\n",
      "Epoch 95/3000\n",
      " - 0s - loss: 0.5083 - mean_squared_error: 0.1615 - acc: 0.8790 - val_loss: 0.5281 - val_mean_squared_error: 0.1709 - val_acc: 0.8577\n",
      "Epoch 96/3000\n",
      " - 0s - loss: 0.5064 - mean_squared_error: 0.1607 - acc: 0.8753 - val_loss: 0.5269 - val_mean_squared_error: 0.1704 - val_acc: 0.8577\n",
      "Epoch 97/3000\n",
      " - 0s - loss: 0.5033 - mean_squared_error: 0.1591 - acc: 0.8826 - val_loss: 0.5260 - val_mean_squared_error: 0.1700 - val_acc: 0.8577\n",
      "Epoch 98/3000\n",
      " - 0s - loss: 0.5036 - mean_squared_error: 0.1594 - acc: 0.8802 - val_loss: 0.5253 - val_mean_squared_error: 0.1696 - val_acc: 0.8577\n",
      "Epoch 99/3000\n",
      " - 0s - loss: 0.5032 - mean_squared_error: 0.1592 - acc: 0.8851 - val_loss: 0.5247 - val_mean_squared_error: 0.1694 - val_acc: 0.8577\n",
      "Epoch 100/3000\n",
      " - 0s - loss: 0.5011 - mean_squared_error: 0.1581 - acc: 0.8802 - val_loss: 0.5244 - val_mean_squared_error: 0.1692 - val_acc: 0.8577\n",
      "Epoch 101/3000\n",
      " - 0s - loss: 0.5023 - mean_squared_error: 0.1589 - acc: 0.8790 - val_loss: 0.5241 - val_mean_squared_error: 0.1691 - val_acc: 0.8577\n",
      "Epoch 102/3000\n",
      " - 0s - loss: 0.5026 - mean_squared_error: 0.1590 - acc: 0.8680 - val_loss: 0.5237 - val_mean_squared_error: 0.1689 - val_acc: 0.8577\n",
      "Epoch 103/3000\n",
      " - 0s - loss: 0.5013 - mean_squared_error: 0.1584 - acc: 0.8790 - val_loss: 0.5231 - val_mean_squared_error: 0.1687 - val_acc: 0.8577\n",
      "Epoch 104/3000\n",
      " - 0s - loss: 0.5008 - mean_squared_error: 0.1582 - acc: 0.8667 - val_loss: 0.5222 - val_mean_squared_error: 0.1683 - val_acc: 0.8577\n",
      "Epoch 105/3000\n",
      " - 0s - loss: 0.4978 - mean_squared_error: 0.1567 - acc: 0.8839 - val_loss: 0.5213 - val_mean_squared_error: 0.1679 - val_acc: 0.8577\n",
      "Epoch 106/3000\n",
      " - 0s - loss: 0.4997 - mean_squared_error: 0.1577 - acc: 0.8790 - val_loss: 0.5201 - val_mean_squared_error: 0.1673 - val_acc: 0.8577\n",
      "Epoch 107/3000\n",
      " - 0s - loss: 0.5014 - mean_squared_error: 0.1585 - acc: 0.8680 - val_loss: 0.5187 - val_mean_squared_error: 0.1668 - val_acc: 0.8613\n",
      "Epoch 108/3000\n",
      " - 0s - loss: 0.4940 - mean_squared_error: 0.1550 - acc: 0.8936 - val_loss: 0.5172 - val_mean_squared_error: 0.1661 - val_acc: 0.8613\n",
      "Epoch 109/3000\n",
      " - 0s - loss: 0.4926 - mean_squared_error: 0.1544 - acc: 0.8851 - val_loss: 0.5155 - val_mean_squared_error: 0.1653 - val_acc: 0.8613\n",
      "Epoch 110/3000\n",
      " - 0s - loss: 0.4943 - mean_squared_error: 0.1555 - acc: 0.8655 - val_loss: 0.5136 - val_mean_squared_error: 0.1645 - val_acc: 0.8650\n",
      "Epoch 111/3000\n",
      " - 0s - loss: 0.4940 - mean_squared_error: 0.1555 - acc: 0.8790 - val_loss: 0.5117 - val_mean_squared_error: 0.1637 - val_acc: 0.8650\n",
      "Epoch 112/3000\n",
      " - 0s - loss: 0.4881 - mean_squared_error: 0.1525 - acc: 0.8900 - val_loss: 0.5100 - val_mean_squared_error: 0.1629 - val_acc: 0.8650\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 113/3000\n",
      " - 0s - loss: 0.4867 - mean_squared_error: 0.1518 - acc: 0.8765 - val_loss: 0.5085 - val_mean_squared_error: 0.1623 - val_acc: 0.8650\n",
      "Epoch 114/3000\n",
      " - 0s - loss: 0.4857 - mean_squared_error: 0.1517 - acc: 0.8765 - val_loss: 0.5071 - val_mean_squared_error: 0.1617 - val_acc: 0.8650\n",
      "Epoch 115/3000\n",
      " - 0s - loss: 0.4818 - mean_squared_error: 0.1500 - acc: 0.8814 - val_loss: 0.5059 - val_mean_squared_error: 0.1612 - val_acc: 0.8650\n",
      "Epoch 116/3000\n",
      " - 0s - loss: 0.4821 - mean_squared_error: 0.1502 - acc: 0.8778 - val_loss: 0.5050 - val_mean_squared_error: 0.1608 - val_acc: 0.8650\n",
      "Epoch 117/3000\n",
      " - 0s - loss: 0.4775 - mean_squared_error: 0.1480 - acc: 0.8839 - val_loss: 0.5042 - val_mean_squared_error: 0.1604 - val_acc: 0.8650\n",
      "Epoch 118/3000\n",
      " - 0s - loss: 0.4776 - mean_squared_error: 0.1481 - acc: 0.8741 - val_loss: 0.5035 - val_mean_squared_error: 0.1601 - val_acc: 0.8650\n",
      "Epoch 119/3000\n",
      " - 0s - loss: 0.4770 - mean_squared_error: 0.1477 - acc: 0.8912 - val_loss: 0.5030 - val_mean_squared_error: 0.1599 - val_acc: 0.8650\n",
      "Epoch 120/3000\n",
      " - 0s - loss: 0.4760 - mean_squared_error: 0.1473 - acc: 0.8912 - val_loss: 0.5027 - val_mean_squared_error: 0.1598 - val_acc: 0.8650\n",
      "Epoch 121/3000\n",
      " - 0s - loss: 0.4741 - mean_squared_error: 0.1466 - acc: 0.8851 - val_loss: 0.5025 - val_mean_squared_error: 0.1597 - val_acc: 0.8613\n",
      "Epoch 122/3000\n",
      " - 0s - loss: 0.4754 - mean_squared_error: 0.1468 - acc: 0.8875 - val_loss: 0.5021 - val_mean_squared_error: 0.1595 - val_acc: 0.8650\n",
      "Epoch 123/3000\n",
      " - 0s - loss: 0.4743 - mean_squared_error: 0.1466 - acc: 0.8900 - val_loss: 0.5016 - val_mean_squared_error: 0.1593 - val_acc: 0.8650\n",
      "Epoch 124/3000\n",
      " - 0s - loss: 0.4803 - mean_squared_error: 0.1495 - acc: 0.8814 - val_loss: 0.5009 - val_mean_squared_error: 0.1590 - val_acc: 0.8650\n",
      "Epoch 125/3000\n",
      " - 0s - loss: 0.4697 - mean_squared_error: 0.1447 - acc: 0.8863 - val_loss: 0.5000 - val_mean_squared_error: 0.1586 - val_acc: 0.8650\n",
      "Epoch 126/3000\n",
      " - 0s - loss: 0.4753 - mean_squared_error: 0.1472 - acc: 0.8839 - val_loss: 0.4990 - val_mean_squared_error: 0.1582 - val_acc: 0.8650\n",
      "Epoch 127/3000\n",
      " - 0s - loss: 0.4709 - mean_squared_error: 0.1451 - acc: 0.8826 - val_loss: 0.4978 - val_mean_squared_error: 0.1577 - val_acc: 0.8650\n",
      "Epoch 128/3000\n",
      " - 0s - loss: 0.4682 - mean_squared_error: 0.1441 - acc: 0.8790 - val_loss: 0.4965 - val_mean_squared_error: 0.1571 - val_acc: 0.8650\n",
      "Epoch 129/3000\n",
      " - 0s - loss: 0.4703 - mean_squared_error: 0.1451 - acc: 0.8826 - val_loss: 0.4950 - val_mean_squared_error: 0.1565 - val_acc: 0.8650\n",
      "Epoch 130/3000\n",
      " - 0s - loss: 0.4702 - mean_squared_error: 0.1449 - acc: 0.8924 - val_loss: 0.4934 - val_mean_squared_error: 0.1558 - val_acc: 0.8650\n",
      "Epoch 131/3000\n",
      " - 0s - loss: 0.4610 - mean_squared_error: 0.1410 - acc: 0.8900 - val_loss: 0.4917 - val_mean_squared_error: 0.1551 - val_acc: 0.8650\n",
      "Epoch 132/3000\n",
      " - 0s - loss: 0.4629 - mean_squared_error: 0.1417 - acc: 0.8900 - val_loss: 0.4902 - val_mean_squared_error: 0.1545 - val_acc: 0.8650\n",
      "Epoch 133/3000\n",
      " - 0s - loss: 0.4626 - mean_squared_error: 0.1417 - acc: 0.8875 - val_loss: 0.4888 - val_mean_squared_error: 0.1539 - val_acc: 0.8650\n",
      "Epoch 134/3000\n",
      " - 0s - loss: 0.4623 - mean_squared_error: 0.1415 - acc: 0.8875 - val_loss: 0.4877 - val_mean_squared_error: 0.1534 - val_acc: 0.8650\n",
      "Epoch 135/3000\n",
      " - 0s - loss: 0.4590 - mean_squared_error: 0.1402 - acc: 0.8924 - val_loss: 0.4866 - val_mean_squared_error: 0.1530 - val_acc: 0.8650\n",
      "Epoch 136/3000\n",
      " - 0s - loss: 0.4600 - mean_squared_error: 0.1407 - acc: 0.8839 - val_loss: 0.4858 - val_mean_squared_error: 0.1526 - val_acc: 0.8650\n",
      "Epoch 137/3000\n",
      " - 0s - loss: 0.4551 - mean_squared_error: 0.1383 - acc: 0.8875 - val_loss: 0.4851 - val_mean_squared_error: 0.1523 - val_acc: 0.8650\n",
      "Epoch 138/3000\n",
      " - 0s - loss: 0.4570 - mean_squared_error: 0.1394 - acc: 0.8863 - val_loss: 0.4845 - val_mean_squared_error: 0.1521 - val_acc: 0.8650\n",
      "Epoch 139/3000\n",
      " - 0s - loss: 0.4506 - mean_squared_error: 0.1364 - acc: 0.8936 - val_loss: 0.4841 - val_mean_squared_error: 0.1519 - val_acc: 0.8650\n",
      "Epoch 140/3000\n",
      " - 0s - loss: 0.4621 - mean_squared_error: 0.1414 - acc: 0.8900 - val_loss: 0.4838 - val_mean_squared_error: 0.1518 - val_acc: 0.8650\n",
      "Epoch 141/3000\n",
      " - 0s - loss: 0.4551 - mean_squared_error: 0.1383 - acc: 0.8936 - val_loss: 0.4836 - val_mean_squared_error: 0.1517 - val_acc: 0.8650\n",
      "Epoch 142/3000\n",
      " - 0s - loss: 0.4528 - mean_squared_error: 0.1373 - acc: 0.8888 - val_loss: 0.4833 - val_mean_squared_error: 0.1516 - val_acc: 0.8650\n",
      "Epoch 143/3000\n",
      " - 0s - loss: 0.4551 - mean_squared_error: 0.1385 - acc: 0.8912 - val_loss: 0.4828 - val_mean_squared_error: 0.1514 - val_acc: 0.8650\n",
      "Epoch 144/3000\n",
      " - 0s - loss: 0.4560 - mean_squared_error: 0.1390 - acc: 0.8863 - val_loss: 0.4822 - val_mean_squared_error: 0.1511 - val_acc: 0.8650\n",
      "Epoch 145/3000\n",
      " - 0s - loss: 0.4520 - mean_squared_error: 0.1371 - acc: 0.8985 - val_loss: 0.4814 - val_mean_squared_error: 0.1508 - val_acc: 0.8650\n",
      "Epoch 146/3000\n",
      " - 0s - loss: 0.4529 - mean_squared_error: 0.1375 - acc: 0.8863 - val_loss: 0.4805 - val_mean_squared_error: 0.1504 - val_acc: 0.8650\n",
      "Epoch 147/3000\n",
      " - 0s - loss: 0.4518 - mean_squared_error: 0.1371 - acc: 0.8900 - val_loss: 0.4795 - val_mean_squared_error: 0.1500 - val_acc: 0.8650\n",
      "Epoch 148/3000\n",
      " - 0s - loss: 0.4491 - mean_squared_error: 0.1355 - acc: 0.8998 - val_loss: 0.4783 - val_mean_squared_error: 0.1495 - val_acc: 0.8686\n",
      "Epoch 149/3000\n",
      " - 0s - loss: 0.4475 - mean_squared_error: 0.1350 - acc: 0.8998 - val_loss: 0.4770 - val_mean_squared_error: 0.1490 - val_acc: 0.8686\n",
      "Epoch 150/3000\n",
      " - 0s - loss: 0.4476 - mean_squared_error: 0.1352 - acc: 0.8924 - val_loss: 0.4756 - val_mean_squared_error: 0.1484 - val_acc: 0.8686\n",
      "Epoch 151/3000\n",
      " - 0s - loss: 0.4428 - mean_squared_error: 0.1331 - acc: 0.8924 - val_loss: 0.4741 - val_mean_squared_error: 0.1478 - val_acc: 0.8686\n",
      "Epoch 152/3000\n",
      " - 0s - loss: 0.4481 - mean_squared_error: 0.1355 - acc: 0.8936 - val_loss: 0.4728 - val_mean_squared_error: 0.1472 - val_acc: 0.8686\n",
      "Epoch 153/3000\n",
      " - 0s - loss: 0.4388 - mean_squared_error: 0.1317 - acc: 0.8961 - val_loss: 0.4716 - val_mean_squared_error: 0.1467 - val_acc: 0.8686\n",
      "Epoch 154/3000\n",
      " - 0s - loss: 0.4374 - mean_squared_error: 0.1313 - acc: 0.8936 - val_loss: 0.4706 - val_mean_squared_error: 0.1463 - val_acc: 0.8686\n",
      "Epoch 155/3000\n",
      " - 0s - loss: 0.4403 - mean_squared_error: 0.1323 - acc: 0.8924 - val_loss: 0.4697 - val_mean_squared_error: 0.1459 - val_acc: 0.8686\n",
      "Epoch 156/3000\n",
      " - 0s - loss: 0.4397 - mean_squared_error: 0.1321 - acc: 0.8949 - val_loss: 0.4689 - val_mean_squared_error: 0.1456 - val_acc: 0.8686\n",
      "Epoch 157/3000\n",
      " - 0s - loss: 0.4385 - mean_squared_error: 0.1312 - acc: 0.8936 - val_loss: 0.4683 - val_mean_squared_error: 0.1454 - val_acc: 0.8686\n",
      "Epoch 158/3000\n",
      " - 0s - loss: 0.4318 - mean_squared_error: 0.1285 - acc: 0.9010 - val_loss: 0.4678 - val_mean_squared_error: 0.1452 - val_acc: 0.8686\n",
      "Epoch 159/3000\n",
      " - 0s - loss: 0.4349 - mean_squared_error: 0.1302 - acc: 0.8936 - val_loss: 0.4675 - val_mean_squared_error: 0.1450 - val_acc: 0.8686\n",
      "Epoch 160/3000\n",
      " - 0s - loss: 0.4382 - mean_squared_error: 0.1315 - acc: 0.8900 - val_loss: 0.4672 - val_mean_squared_error: 0.1449 - val_acc: 0.8686\n",
      "Epoch 161/3000\n",
      " - 0s - loss: 0.4350 - mean_squared_error: 0.1302 - acc: 0.8973 - val_loss: 0.4671 - val_mean_squared_error: 0.1449 - val_acc: 0.8686\n",
      "Epoch 162/3000\n",
      " - 0s - loss: 0.4398 - mean_squared_error: 0.1322 - acc: 0.8888 - val_loss: 0.4668 - val_mean_squared_error: 0.1447 - val_acc: 0.8686\n",
      "Epoch 163/3000\n",
      " - 0s - loss: 0.4372 - mean_squared_error: 0.1313 - acc: 0.8863 - val_loss: 0.4663 - val_mean_squared_error: 0.1446 - val_acc: 0.8686\n",
      "Epoch 164/3000\n",
      " - 0s - loss: 0.4306 - mean_squared_error: 0.1280 - acc: 0.8998 - val_loss: 0.4658 - val_mean_squared_error: 0.1443 - val_acc: 0.8686\n",
      "Epoch 165/3000\n",
      " - 0s - loss: 0.4339 - mean_squared_error: 0.1297 - acc: 0.9022 - val_loss: 0.4651 - val_mean_squared_error: 0.1441 - val_acc: 0.8686\n",
      "Epoch 166/3000\n",
      " - 0s - loss: 0.4283 - mean_squared_error: 0.1273 - acc: 0.8973 - val_loss: 0.4643 - val_mean_squared_error: 0.1437 - val_acc: 0.8686\n",
      "Epoch 167/3000\n",
      " - 0s - loss: 0.4313 - mean_squared_error: 0.1286 - acc: 0.9010 - val_loss: 0.4634 - val_mean_squared_error: 0.1434 - val_acc: 0.8686\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 168/3000\n",
      " - 0s - loss: 0.4274 - mean_squared_error: 0.1266 - acc: 0.9071 - val_loss: 0.4624 - val_mean_squared_error: 0.1430 - val_acc: 0.8686\n",
      "Epoch 169/3000\n",
      " - 0s - loss: 0.4281 - mean_squared_error: 0.1271 - acc: 0.9022 - val_loss: 0.4612 - val_mean_squared_error: 0.1425 - val_acc: 0.8686\n",
      "Epoch 170/3000\n",
      " - 0s - loss: 0.4283 - mean_squared_error: 0.1274 - acc: 0.8875 - val_loss: 0.4600 - val_mean_squared_error: 0.1420 - val_acc: 0.8686\n",
      "Epoch 171/3000\n",
      " - 0s - loss: 0.4271 - mean_squared_error: 0.1271 - acc: 0.8924 - val_loss: 0.4586 - val_mean_squared_error: 0.1415 - val_acc: 0.8686\n",
      "Epoch 172/3000\n",
      " - 0s - loss: 0.4263 - mean_squared_error: 0.1268 - acc: 0.8888 - val_loss: 0.4575 - val_mean_squared_error: 0.1410 - val_acc: 0.8686\n",
      "Epoch 173/3000\n",
      " - 0s - loss: 0.4233 - mean_squared_error: 0.1254 - acc: 0.8998 - val_loss: 0.4564 - val_mean_squared_error: 0.1406 - val_acc: 0.8686\n",
      "Epoch 174/3000\n",
      " - 0s - loss: 0.4225 - mean_squared_error: 0.1253 - acc: 0.8973 - val_loss: 0.4555 - val_mean_squared_error: 0.1402 - val_acc: 0.8723\n",
      "Epoch 175/3000\n",
      " - 0s - loss: 0.4206 - mean_squared_error: 0.1242 - acc: 0.9022 - val_loss: 0.4547 - val_mean_squared_error: 0.1399 - val_acc: 0.8723\n",
      "Epoch 176/3000\n",
      " - 0s - loss: 0.4159 - mean_squared_error: 0.1225 - acc: 0.8961 - val_loss: 0.4540 - val_mean_squared_error: 0.1396 - val_acc: 0.8723\n",
      "Epoch 177/3000\n",
      " - 0s - loss: 0.4167 - mean_squared_error: 0.1226 - acc: 0.9046 - val_loss: 0.4535 - val_mean_squared_error: 0.1394 - val_acc: 0.8723\n",
      "Epoch 178/3000\n",
      " - 0s - loss: 0.4204 - mean_squared_error: 0.1242 - acc: 0.8961 - val_loss: 0.4530 - val_mean_squared_error: 0.1392 - val_acc: 0.8723\n",
      "Epoch 179/3000\n",
      " - 0s - loss: 0.4131 - mean_squared_error: 0.1212 - acc: 0.8924 - val_loss: 0.4527 - val_mean_squared_error: 0.1391 - val_acc: 0.8723\n",
      "Epoch 180/3000\n",
      " - 0s - loss: 0.4189 - mean_squared_error: 0.1236 - acc: 0.8985 - val_loss: 0.4525 - val_mean_squared_error: 0.1390 - val_acc: 0.8723\n",
      "Epoch 181/3000\n",
      " - 0s - loss: 0.4148 - mean_squared_error: 0.1217 - acc: 0.9059 - val_loss: 0.4523 - val_mean_squared_error: 0.1389 - val_acc: 0.8723\n",
      "Epoch 182/3000\n",
      " - 0s - loss: 0.4159 - mean_squared_error: 0.1222 - acc: 0.9022 - val_loss: 0.4520 - val_mean_squared_error: 0.1388 - val_acc: 0.8723\n",
      "Epoch 183/3000\n",
      " - 0s - loss: 0.4185 - mean_squared_error: 0.1236 - acc: 0.8985 - val_loss: 0.4517 - val_mean_squared_error: 0.1387 - val_acc: 0.8723\n",
      "Epoch 184/3000\n",
      " - 0s - loss: 0.4139 - mean_squared_error: 0.1211 - acc: 0.9095 - val_loss: 0.4512 - val_mean_squared_error: 0.1385 - val_acc: 0.8686\n",
      "Epoch 185/3000\n",
      " - 0s - loss: 0.4119 - mean_squared_error: 0.1209 - acc: 0.9034 - val_loss: 0.4506 - val_mean_squared_error: 0.1382 - val_acc: 0.8686\n",
      "Epoch 186/3000\n",
      " - 0s - loss: 0.4138 - mean_squared_error: 0.1218 - acc: 0.8924 - val_loss: 0.4499 - val_mean_squared_error: 0.1380 - val_acc: 0.8686\n",
      "Epoch 187/3000\n",
      " - 0s - loss: 0.4142 - mean_squared_error: 0.1217 - acc: 0.8912 - val_loss: 0.4490 - val_mean_squared_error: 0.1376 - val_acc: 0.8686\n",
      "Epoch 188/3000\n",
      " - 0s - loss: 0.4092 - mean_squared_error: 0.1198 - acc: 0.9034 - val_loss: 0.4481 - val_mean_squared_error: 0.1373 - val_acc: 0.8686\n",
      "Epoch 189/3000\n",
      " - 0s - loss: 0.4085 - mean_squared_error: 0.1197 - acc: 0.8924 - val_loss: 0.4471 - val_mean_squared_error: 0.1369 - val_acc: 0.8686\n",
      "Epoch 190/3000\n",
      " - 0s - loss: 0.4060 - mean_squared_error: 0.1183 - acc: 0.9108 - val_loss: 0.4460 - val_mean_squared_error: 0.1364 - val_acc: 0.8686\n",
      "Epoch 191/3000\n",
      " - 0s - loss: 0.4062 - mean_squared_error: 0.1187 - acc: 0.8973 - val_loss: 0.4448 - val_mean_squared_error: 0.1360 - val_acc: 0.8686\n",
      "Epoch 192/3000\n",
      " - 0s - loss: 0.4065 - mean_squared_error: 0.1187 - acc: 0.9095 - val_loss: 0.4437 - val_mean_squared_error: 0.1356 - val_acc: 0.8686\n",
      "Epoch 193/3000\n",
      " - 0s - loss: 0.4052 - mean_squared_error: 0.1181 - acc: 0.9034 - val_loss: 0.4428 - val_mean_squared_error: 0.1352 - val_acc: 0.8686\n",
      "Epoch 194/3000\n",
      " - 0s - loss: 0.4052 - mean_squared_error: 0.1182 - acc: 0.8985 - val_loss: 0.4420 - val_mean_squared_error: 0.1349 - val_acc: 0.8686\n",
      "Epoch 195/3000\n",
      " - 0s - loss: 0.4040 - mean_squared_error: 0.1176 - acc: 0.9083 - val_loss: 0.4413 - val_mean_squared_error: 0.1346 - val_acc: 0.8686\n",
      "Epoch 196/3000\n",
      " - 0s - loss: 0.4003 - mean_squared_error: 0.1160 - acc: 0.9059 - val_loss: 0.4407 - val_mean_squared_error: 0.1344 - val_acc: 0.8686\n",
      "Epoch 197/3000\n",
      " - 0s - loss: 0.4014 - mean_squared_error: 0.1166 - acc: 0.9120 - val_loss: 0.4402 - val_mean_squared_error: 0.1342 - val_acc: 0.8686\n",
      "Epoch 198/3000\n",
      " - 0s - loss: 0.4021 - mean_squared_error: 0.1168 - acc: 0.9034 - val_loss: 0.4398 - val_mean_squared_error: 0.1340 - val_acc: 0.8686\n",
      "Epoch 199/3000\n",
      " - 0s - loss: 0.4022 - mean_squared_error: 0.1168 - acc: 0.9071 - val_loss: 0.4395 - val_mean_squared_error: 0.1339 - val_acc: 0.8686\n",
      "Epoch 200/3000\n",
      " - 0s - loss: 0.3977 - mean_squared_error: 0.1152 - acc: 0.8985 - val_loss: 0.4393 - val_mean_squared_error: 0.1338 - val_acc: 0.8686\n",
      "Epoch 201/3000\n",
      " - 0s - loss: 0.3993 - mean_squared_error: 0.1158 - acc: 0.9046 - val_loss: 0.4391 - val_mean_squared_error: 0.1338 - val_acc: 0.8686\n",
      "Epoch 202/3000\n",
      " - 0s - loss: 0.3992 - mean_squared_error: 0.1157 - acc: 0.9022 - val_loss: 0.4389 - val_mean_squared_error: 0.1337 - val_acc: 0.8686\n",
      "Epoch 203/3000\n",
      " - 0s - loss: 0.3997 - mean_squared_error: 0.1160 - acc: 0.8973 - val_loss: 0.4386 - val_mean_squared_error: 0.1336 - val_acc: 0.8686\n",
      "Epoch 204/3000\n",
      " - 0s - loss: 0.3968 - mean_squared_error: 0.1149 - acc: 0.9108 - val_loss: 0.4381 - val_mean_squared_error: 0.1334 - val_acc: 0.8686\n",
      "Epoch 205/3000\n",
      " - 0s - loss: 0.3995 - mean_squared_error: 0.1156 - acc: 0.9108 - val_loss: 0.4376 - val_mean_squared_error: 0.1332 - val_acc: 0.8686\n",
      "Epoch 206/3000\n",
      " - 0s - loss: 0.3954 - mean_squared_error: 0.1143 - acc: 0.9108 - val_loss: 0.4370 - val_mean_squared_error: 0.1329 - val_acc: 0.8686\n",
      "Epoch 207/3000\n",
      " - 0s - loss: 0.3951 - mean_squared_error: 0.1141 - acc: 0.9034 - val_loss: 0.4362 - val_mean_squared_error: 0.1327 - val_acc: 0.8686\n",
      "Epoch 208/3000\n",
      " - 0s - loss: 0.3956 - mean_squared_error: 0.1142 - acc: 0.9108 - val_loss: 0.4354 - val_mean_squared_error: 0.1323 - val_acc: 0.8686\n",
      "Epoch 209/3000\n",
      " - 0s - loss: 0.3933 - mean_squared_error: 0.1134 - acc: 0.9010 - val_loss: 0.4345 - val_mean_squared_error: 0.1320 - val_acc: 0.8686\n",
      "Epoch 210/3000\n",
      " - 0s - loss: 0.3927 - mean_squared_error: 0.1131 - acc: 0.9156 - val_loss: 0.4335 - val_mean_squared_error: 0.1316 - val_acc: 0.8686\n",
      "Epoch 211/3000\n",
      " - 0s - loss: 0.3908 - mean_squared_error: 0.1125 - acc: 0.8998 - val_loss: 0.4325 - val_mean_squared_error: 0.1312 - val_acc: 0.8686\n",
      "Epoch 212/3000\n",
      " - 0s - loss: 0.3936 - mean_squared_error: 0.1137 - acc: 0.9095 - val_loss: 0.4315 - val_mean_squared_error: 0.1309 - val_acc: 0.8686\n",
      "Epoch 213/3000\n",
      " - 0s - loss: 0.3855 - mean_squared_error: 0.1103 - acc: 0.9046 - val_loss: 0.4307 - val_mean_squared_error: 0.1305 - val_acc: 0.8686\n",
      "Epoch 214/3000\n",
      " - 0s - loss: 0.3865 - mean_squared_error: 0.1109 - acc: 0.9083 - val_loss: 0.4300 - val_mean_squared_error: 0.1302 - val_acc: 0.8686\n",
      "Epoch 215/3000\n",
      " - 0s - loss: 0.3866 - mean_squared_error: 0.1106 - acc: 0.9132 - val_loss: 0.4293 - val_mean_squared_error: 0.1300 - val_acc: 0.8686\n",
      "Epoch 216/3000\n",
      " - 0s - loss: 0.3878 - mean_squared_error: 0.1113 - acc: 0.9034 - val_loss: 0.4288 - val_mean_squared_error: 0.1298 - val_acc: 0.8686\n",
      "Epoch 217/3000\n",
      " - 0s - loss: 0.3835 - mean_squared_error: 0.1097 - acc: 0.9144 - val_loss: 0.4283 - val_mean_squared_error: 0.1296 - val_acc: 0.8686\n",
      "Epoch 218/3000\n",
      " - 0s - loss: 0.3845 - mean_squared_error: 0.1100 - acc: 0.9108 - val_loss: 0.4280 - val_mean_squared_error: 0.1295 - val_acc: 0.8686\n",
      "Epoch 219/3000\n",
      " - 0s - loss: 0.3876 - mean_squared_error: 0.1114 - acc: 0.9071 - val_loss: 0.4277 - val_mean_squared_error: 0.1294 - val_acc: 0.8686\n",
      "Epoch 220/3000\n",
      " - 0s - loss: 0.3850 - mean_squared_error: 0.1106 - acc: 0.9046 - val_loss: 0.4275 - val_mean_squared_error: 0.1293 - val_acc: 0.8686\n",
      "Epoch 221/3000\n",
      " - 0s - loss: 0.3872 - mean_squared_error: 0.1110 - acc: 0.9120 - val_loss: 0.4274 - val_mean_squared_error: 0.1293 - val_acc: 0.8686\n",
      "Epoch 222/3000\n",
      " - 0s - loss: 0.3827 - mean_squared_error: 0.1090 - acc: 0.9132 - val_loss: 0.4272 - val_mean_squared_error: 0.1292 - val_acc: 0.8686\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 223/3000\n",
      " - 0s - loss: 0.3871 - mean_squared_error: 0.1117 - acc: 0.9010 - val_loss: 0.4269 - val_mean_squared_error: 0.1291 - val_acc: 0.8686\n",
      "Epoch 224/3000\n",
      " - 0s - loss: 0.3823 - mean_squared_error: 0.1094 - acc: 0.9059 - val_loss: 0.4265 - val_mean_squared_error: 0.1289 - val_acc: 0.8686\n",
      "Epoch 225/3000\n",
      " - 0s - loss: 0.3841 - mean_squared_error: 0.1097 - acc: 0.9095 - val_loss: 0.4260 - val_mean_squared_error: 0.1288 - val_acc: 0.8686\n",
      "Epoch 226/3000\n",
      " - 0s - loss: 0.3811 - mean_squared_error: 0.1087 - acc: 0.9144 - val_loss: 0.4254 - val_mean_squared_error: 0.1285 - val_acc: 0.8686\n",
      "Epoch 227/3000\n",
      " - 0s - loss: 0.3807 - mean_squared_error: 0.1086 - acc: 0.9156 - val_loss: 0.4248 - val_mean_squared_error: 0.1283 - val_acc: 0.8686\n",
      "Epoch 228/3000\n",
      " - 0s - loss: 0.3815 - mean_squared_error: 0.1090 - acc: 0.9095 - val_loss: 0.4240 - val_mean_squared_error: 0.1280 - val_acc: 0.8686\n",
      "Epoch 229/3000\n",
      " - 0s - loss: 0.3758 - mean_squared_error: 0.1066 - acc: 0.9095 - val_loss: 0.4232 - val_mean_squared_error: 0.1277 - val_acc: 0.8686\n",
      "Epoch 230/3000\n",
      " - 0s - loss: 0.3793 - mean_squared_error: 0.1085 - acc: 0.9120 - val_loss: 0.4223 - val_mean_squared_error: 0.1274 - val_acc: 0.8686\n",
      "Epoch 231/3000\n",
      " - 0s - loss: 0.3772 - mean_squared_error: 0.1074 - acc: 0.9108 - val_loss: 0.4214 - val_mean_squared_error: 0.1270 - val_acc: 0.8686\n",
      "Epoch 232/3000\n",
      " - 0s - loss: 0.3745 - mean_squared_error: 0.1065 - acc: 0.9022 - val_loss: 0.4205 - val_mean_squared_error: 0.1267 - val_acc: 0.8686\n",
      "Epoch 233/3000\n",
      " - 0s - loss: 0.3766 - mean_squared_error: 0.1071 - acc: 0.9156 - val_loss: 0.4198 - val_mean_squared_error: 0.1264 - val_acc: 0.8686\n",
      "Epoch 234/3000\n",
      " - 0s - loss: 0.3738 - mean_squared_error: 0.1058 - acc: 0.9108 - val_loss: 0.4191 - val_mean_squared_error: 0.1262 - val_acc: 0.8686\n",
      "Epoch 235/3000\n",
      " - 0s - loss: 0.3718 - mean_squared_error: 0.1051 - acc: 0.9144 - val_loss: 0.4185 - val_mean_squared_error: 0.1259 - val_acc: 0.8686\n",
      "Epoch 236/3000\n",
      " - 0s - loss: 0.3716 - mean_squared_error: 0.1052 - acc: 0.9156 - val_loss: 0.4180 - val_mean_squared_error: 0.1258 - val_acc: 0.8686\n",
      "Epoch 237/3000\n",
      " - 0s - loss: 0.3732 - mean_squared_error: 0.1059 - acc: 0.9046 - val_loss: 0.4176 - val_mean_squared_error: 0.1256 - val_acc: 0.8686\n",
      "Epoch 238/3000\n",
      " - 0s - loss: 0.3671 - mean_squared_error: 0.1035 - acc: 0.9132 - val_loss: 0.4173 - val_mean_squared_error: 0.1255 - val_acc: 0.8686\n",
      "Epoch 239/3000\n",
      " - 0s - loss: 0.3721 - mean_squared_error: 0.1057 - acc: 0.9083 - val_loss: 0.4171 - val_mean_squared_error: 0.1254 - val_acc: 0.8686\n",
      "Epoch 240/3000\n",
      " - 0s - loss: 0.3741 - mean_squared_error: 0.1067 - acc: 0.9059 - val_loss: 0.4169 - val_mean_squared_error: 0.1253 - val_acc: 0.8686\n",
      "Epoch 241/3000\n",
      " - 0s - loss: 0.3712 - mean_squared_error: 0.1053 - acc: 0.9083 - val_loss: 0.4168 - val_mean_squared_error: 0.1253 - val_acc: 0.8686\n",
      "Epoch 242/3000\n",
      " - 0s - loss: 0.3722 - mean_squared_error: 0.1056 - acc: 0.9022 - val_loss: 0.4166 - val_mean_squared_error: 0.1252 - val_acc: 0.8686\n",
      "Epoch 243/3000\n",
      " - 0s - loss: 0.3719 - mean_squared_error: 0.1054 - acc: 0.9059 - val_loss: 0.4163 - val_mean_squared_error: 0.1251 - val_acc: 0.8686\n",
      "Epoch 244/3000\n",
      " - 0s - loss: 0.3704 - mean_squared_error: 0.1049 - acc: 0.9132 - val_loss: 0.4160 - val_mean_squared_error: 0.1250 - val_acc: 0.8686\n",
      "Epoch 245/3000\n",
      " - 0s - loss: 0.3664 - mean_squared_error: 0.1029 - acc: 0.9181 - val_loss: 0.4155 - val_mean_squared_error: 0.1248 - val_acc: 0.8686\n",
      "Epoch 246/3000\n",
      " - 0s - loss: 0.3704 - mean_squared_error: 0.1047 - acc: 0.9132 - val_loss: 0.4150 - val_mean_squared_error: 0.1247 - val_acc: 0.8686\n",
      "Epoch 247/3000\n",
      " - 0s - loss: 0.3672 - mean_squared_error: 0.1034 - acc: 0.9169 - val_loss: 0.4144 - val_mean_squared_error: 0.1244 - val_acc: 0.8686\n",
      "Epoch 248/3000\n",
      " - 0s - loss: 0.3691 - mean_squared_error: 0.1044 - acc: 0.9108 - val_loss: 0.4138 - val_mean_squared_error: 0.1242 - val_acc: 0.8686\n",
      "Epoch 249/3000\n",
      " - 0s - loss: 0.3686 - mean_squared_error: 0.1041 - acc: 0.9071 - val_loss: 0.4130 - val_mean_squared_error: 0.1239 - val_acc: 0.8686\n",
      "Epoch 250/3000\n",
      " - 0s - loss: 0.3671 - mean_squared_error: 0.1035 - acc: 0.9193 - val_loss: 0.4122 - val_mean_squared_error: 0.1236 - val_acc: 0.8686\n",
      "Epoch 251/3000\n",
      " - 0s - loss: 0.3650 - mean_squared_error: 0.1028 - acc: 0.9132 - val_loss: 0.4114 - val_mean_squared_error: 0.1233 - val_acc: 0.8686\n",
      "Epoch 252/3000\n",
      " - 0s - loss: 0.3622 - mean_squared_error: 0.1017 - acc: 0.9120 - val_loss: 0.4106 - val_mean_squared_error: 0.1230 - val_acc: 0.8686\n",
      "Epoch 253/3000\n",
      " - 0s - loss: 0.3647 - mean_squared_error: 0.1024 - acc: 0.9095 - val_loss: 0.4099 - val_mean_squared_error: 0.1228 - val_acc: 0.8686\n",
      "Epoch 254/3000\n",
      " - 0s - loss: 0.3627 - mean_squared_error: 0.1020 - acc: 0.9083 - val_loss: 0.4093 - val_mean_squared_error: 0.1225 - val_acc: 0.8686\n",
      "Epoch 255/3000\n",
      " - 0s - loss: 0.3623 - mean_squared_error: 0.1015 - acc: 0.9108 - val_loss: 0.4088 - val_mean_squared_error: 0.1224 - val_acc: 0.8686\n",
      "Epoch 256/3000\n",
      " - 0s - loss: 0.3576 - mean_squared_error: 0.1002 - acc: 0.9181 - val_loss: 0.4083 - val_mean_squared_error: 0.1222 - val_acc: 0.8686\n",
      "Epoch 257/3000\n",
      " - 0s - loss: 0.3590 - mean_squared_error: 0.1005 - acc: 0.9108 - val_loss: 0.4080 - val_mean_squared_error: 0.1221 - val_acc: 0.8686\n",
      "Epoch 258/3000\n",
      " - 0s - loss: 0.3590 - mean_squared_error: 0.1006 - acc: 0.9169 - val_loss: 0.4077 - val_mean_squared_error: 0.1220 - val_acc: 0.8686\n",
      "Epoch 259/3000\n",
      " - 0s - loss: 0.3530 - mean_squared_error: 0.0981 - acc: 0.9108 - val_loss: 0.4075 - val_mean_squared_error: 0.1219 - val_acc: 0.8686\n",
      "Epoch 260/3000\n",
      " - 0s - loss: 0.3574 - mean_squared_error: 0.1002 - acc: 0.9059 - val_loss: 0.4073 - val_mean_squared_error: 0.1218 - val_acc: 0.8686\n",
      "Epoch 261/3000\n",
      " - 0s - loss: 0.3523 - mean_squared_error: 0.0976 - acc: 0.9230 - val_loss: 0.4072 - val_mean_squared_error: 0.1218 - val_acc: 0.8686\n",
      "Epoch 262/3000\n",
      " - 0s - loss: 0.3579 - mean_squared_error: 0.1001 - acc: 0.9046 - val_loss: 0.4070 - val_mean_squared_error: 0.1217 - val_acc: 0.8686\n",
      "Epoch 263/3000\n",
      " - 0s - loss: 0.3548 - mean_squared_error: 0.0990 - acc: 0.9181 - val_loss: 0.4068 - val_mean_squared_error: 0.1216 - val_acc: 0.8686\n",
      "Epoch 264/3000\n",
      " - 0s - loss: 0.3568 - mean_squared_error: 0.0998 - acc: 0.9181 - val_loss: 0.4065 - val_mean_squared_error: 0.1215 - val_acc: 0.8686\n",
      "Epoch 265/3000\n",
      " - 0s - loss: 0.3552 - mean_squared_error: 0.0993 - acc: 0.9181 - val_loss: 0.4061 - val_mean_squared_error: 0.1214 - val_acc: 0.8686\n",
      "Epoch 266/3000\n",
      " - 0s - loss: 0.3565 - mean_squared_error: 0.0996 - acc: 0.9144 - val_loss: 0.4056 - val_mean_squared_error: 0.1212 - val_acc: 0.8686\n",
      "Epoch 267/3000\n",
      " - 0s - loss: 0.3520 - mean_squared_error: 0.0979 - acc: 0.9108 - val_loss: 0.4051 - val_mean_squared_error: 0.1210 - val_acc: 0.8686\n",
      "Epoch 268/3000\n",
      " - 0s - loss: 0.3537 - mean_squared_error: 0.0987 - acc: 0.9120 - val_loss: 0.4044 - val_mean_squared_error: 0.1208 - val_acc: 0.8686\n",
      "Epoch 269/3000\n",
      " - 0s - loss: 0.3503 - mean_squared_error: 0.0973 - acc: 0.9193 - val_loss: 0.4038 - val_mean_squared_error: 0.1205 - val_acc: 0.8650\n",
      "Epoch 270/3000\n",
      " - 0s - loss: 0.3579 - mean_squared_error: 0.1005 - acc: 0.9083 - val_loss: 0.4030 - val_mean_squared_error: 0.1203 - val_acc: 0.8650\n",
      "Epoch 271/3000\n",
      " - 0s - loss: 0.3521 - mean_squared_error: 0.0982 - acc: 0.9120 - val_loss: 0.4022 - val_mean_squared_error: 0.1200 - val_acc: 0.8650\n",
      "Epoch 272/3000\n",
      " - 0s - loss: 0.3472 - mean_squared_error: 0.0962 - acc: 0.9132 - val_loss: 0.4015 - val_mean_squared_error: 0.1197 - val_acc: 0.8650\n",
      "Epoch 273/3000\n",
      " - 0s - loss: 0.3529 - mean_squared_error: 0.0983 - acc: 0.9132 - val_loss: 0.4009 - val_mean_squared_error: 0.1195 - val_acc: 0.8650\n",
      "Epoch 274/3000\n",
      " - 0s - loss: 0.3463 - mean_squared_error: 0.0955 - acc: 0.9181 - val_loss: 0.4004 - val_mean_squared_error: 0.1193 - val_acc: 0.8650\n",
      "Epoch 275/3000\n",
      " - 0s - loss: 0.3461 - mean_squared_error: 0.0958 - acc: 0.9144 - val_loss: 0.3999 - val_mean_squared_error: 0.1192 - val_acc: 0.8650\n",
      "Epoch 276/3000\n",
      " - 0s - loss: 0.3469 - mean_squared_error: 0.0963 - acc: 0.9132 - val_loss: 0.3995 - val_mean_squared_error: 0.1190 - val_acc: 0.8650\n",
      "Epoch 277/3000\n",
      " - 0s - loss: 0.3453 - mean_squared_error: 0.0952 - acc: 0.9169 - val_loss: 0.3991 - val_mean_squared_error: 0.1189 - val_acc: 0.8650\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 278/3000\n",
      " - 0s - loss: 0.3494 - mean_squared_error: 0.0971 - acc: 0.9169 - val_loss: 0.3989 - val_mean_squared_error: 0.1188 - val_acc: 0.8650\n",
      "Epoch 279/3000\n",
      " - 0s - loss: 0.3447 - mean_squared_error: 0.0950 - acc: 0.9193 - val_loss: 0.3987 - val_mean_squared_error: 0.1187 - val_acc: 0.8650\n",
      "Epoch 280/3000\n",
      " - 0s - loss: 0.3432 - mean_squared_error: 0.0950 - acc: 0.9120 - val_loss: 0.3985 - val_mean_squared_error: 0.1187 - val_acc: 0.8650\n",
      "Epoch 281/3000\n",
      " - 0s - loss: 0.3408 - mean_squared_error: 0.0937 - acc: 0.9242 - val_loss: 0.3984 - val_mean_squared_error: 0.1186 - val_acc: 0.8650\n",
      "Epoch 282/3000\n",
      " - 0s - loss: 0.3446 - mean_squared_error: 0.0952 - acc: 0.9132 - val_loss: 0.3983 - val_mean_squared_error: 0.1186 - val_acc: 0.8650\n",
      "Epoch 283/3000\n",
      " - 0s - loss: 0.3494 - mean_squared_error: 0.0974 - acc: 0.9059 - val_loss: 0.3981 - val_mean_squared_error: 0.1185 - val_acc: 0.8650\n",
      "Epoch 284/3000\n",
      " - 0s - loss: 0.3446 - mean_squared_error: 0.0951 - acc: 0.9156 - val_loss: 0.3978 - val_mean_squared_error: 0.1184 - val_acc: 0.8650\n",
      "Epoch 285/3000\n",
      " - 0s - loss: 0.3454 - mean_squared_error: 0.0954 - acc: 0.9279 - val_loss: 0.3974 - val_mean_squared_error: 0.1183 - val_acc: 0.8650\n",
      "Epoch 286/3000\n",
      " - 0s - loss: 0.3492 - mean_squared_error: 0.0976 - acc: 0.9083 - val_loss: 0.3969 - val_mean_squared_error: 0.1181 - val_acc: 0.8650\n",
      "Epoch 287/3000\n",
      " - 0s - loss: 0.3422 - mean_squared_error: 0.0949 - acc: 0.9095 - val_loss: 0.3965 - val_mean_squared_error: 0.1179 - val_acc: 0.8650\n",
      "Epoch 288/3000\n",
      " - 0s - loss: 0.3429 - mean_squared_error: 0.0946 - acc: 0.9120 - val_loss: 0.3959 - val_mean_squared_error: 0.1177 - val_acc: 0.8650\n",
      "Epoch 289/3000\n",
      " - 0s - loss: 0.3391 - mean_squared_error: 0.0932 - acc: 0.9181 - val_loss: 0.3953 - val_mean_squared_error: 0.1175 - val_acc: 0.8650\n",
      "Epoch 290/3000\n",
      " - 0s - loss: 0.3384 - mean_squared_error: 0.0927 - acc: 0.9156 - val_loss: 0.3946 - val_mean_squared_error: 0.1173 - val_acc: 0.8650\n",
      "Epoch 291/3000\n",
      " - 0s - loss: 0.3402 - mean_squared_error: 0.0938 - acc: 0.9120 - val_loss: 0.3939 - val_mean_squared_error: 0.1170 - val_acc: 0.8650\n",
      "Epoch 292/3000\n",
      " - 0s - loss: 0.3385 - mean_squared_error: 0.0927 - acc: 0.9169 - val_loss: 0.3932 - val_mean_squared_error: 0.1168 - val_acc: 0.8650\n",
      "Epoch 293/3000\n",
      " - 0s - loss: 0.3390 - mean_squared_error: 0.0936 - acc: 0.9193 - val_loss: 0.3927 - val_mean_squared_error: 0.1166 - val_acc: 0.8650\n",
      "Epoch 294/3000\n",
      " - 0s - loss: 0.3372 - mean_squared_error: 0.0927 - acc: 0.9230 - val_loss: 0.3922 - val_mean_squared_error: 0.1164 - val_acc: 0.8650\n",
      "Epoch 295/3000\n",
      " - 0s - loss: 0.3361 - mean_squared_error: 0.0922 - acc: 0.9205 - val_loss: 0.3917 - val_mean_squared_error: 0.1163 - val_acc: 0.8650\n",
      "Epoch 296/3000\n",
      " - 0s - loss: 0.3345 - mean_squared_error: 0.0916 - acc: 0.9181 - val_loss: 0.3914 - val_mean_squared_error: 0.1162 - val_acc: 0.8650\n",
      "Epoch 297/3000\n",
      " - 0s - loss: 0.3365 - mean_squared_error: 0.0926 - acc: 0.9108 - val_loss: 0.3911 - val_mean_squared_error: 0.1161 - val_acc: 0.8650\n",
      "Epoch 298/3000\n",
      " - 0s - loss: 0.3325 - mean_squared_error: 0.0905 - acc: 0.9254 - val_loss: 0.3908 - val_mean_squared_error: 0.1160 - val_acc: 0.8650\n",
      "Epoch 299/3000\n",
      " - 0s - loss: 0.3311 - mean_squared_error: 0.0901 - acc: 0.9156 - val_loss: 0.3906 - val_mean_squared_error: 0.1159 - val_acc: 0.8650\n",
      "Epoch 300/3000\n",
      " - 0s - loss: 0.3321 - mean_squared_error: 0.0906 - acc: 0.9279 - val_loss: 0.3905 - val_mean_squared_error: 0.1159 - val_acc: 0.8650\n",
      "Epoch 301/3000\n",
      " - 0s - loss: 0.3300 - mean_squared_error: 0.0900 - acc: 0.9242 - val_loss: 0.3904 - val_mean_squared_error: 0.1158 - val_acc: 0.8650\n",
      "Epoch 302/3000\n",
      " - 0s - loss: 0.3312 - mean_squared_error: 0.0898 - acc: 0.9230 - val_loss: 0.3903 - val_mean_squared_error: 0.1158 - val_acc: 0.8650\n",
      "Epoch 303/3000\n",
      " - 0s - loss: 0.3300 - mean_squared_error: 0.0898 - acc: 0.9218 - val_loss: 0.3901 - val_mean_squared_error: 0.1157 - val_acc: 0.8650\n",
      "Epoch 304/3000\n",
      " - 0s - loss: 0.3287 - mean_squared_error: 0.0895 - acc: 0.9230 - val_loss: 0.3898 - val_mean_squared_error: 0.1156 - val_acc: 0.8650\n",
      "Epoch 305/3000\n",
      " - 0s - loss: 0.3280 - mean_squared_error: 0.0892 - acc: 0.9181 - val_loss: 0.3895 - val_mean_squared_error: 0.1155 - val_acc: 0.8650\n",
      "Epoch 306/3000\n",
      " - 0s - loss: 0.3296 - mean_squared_error: 0.0898 - acc: 0.9279 - val_loss: 0.3891 - val_mean_squared_error: 0.1154 - val_acc: 0.8650\n",
      "Epoch 307/3000\n",
      " - 0s - loss: 0.3297 - mean_squared_error: 0.0901 - acc: 0.9169 - val_loss: 0.3887 - val_mean_squared_error: 0.1152 - val_acc: 0.8650\n",
      "Epoch 308/3000\n",
      " - 0s - loss: 0.3301 - mean_squared_error: 0.0897 - acc: 0.9156 - val_loss: 0.3881 - val_mean_squared_error: 0.1151 - val_acc: 0.8650\n",
      "Epoch 309/3000\n",
      " - 0s - loss: 0.3300 - mean_squared_error: 0.0903 - acc: 0.9169 - val_loss: 0.3876 - val_mean_squared_error: 0.1149 - val_acc: 0.8650\n",
      "Epoch 310/3000\n",
      " - 0s - loss: 0.3298 - mean_squared_error: 0.0898 - acc: 0.9218 - val_loss: 0.3870 - val_mean_squared_error: 0.1146 - val_acc: 0.8650\n",
      "Epoch 311/3000\n",
      " - 0s - loss: 0.3289 - mean_squared_error: 0.0896 - acc: 0.9169 - val_loss: 0.3863 - val_mean_squared_error: 0.1144 - val_acc: 0.8650\n",
      "Epoch 312/3000\n",
      " - 0s - loss: 0.3266 - mean_squared_error: 0.0887 - acc: 0.9230 - val_loss: 0.3857 - val_mean_squared_error: 0.1142 - val_acc: 0.8650\n",
      "Epoch 313/3000\n",
      " - 0s - loss: 0.3249 - mean_squared_error: 0.0880 - acc: 0.9193 - val_loss: 0.3852 - val_mean_squared_error: 0.1141 - val_acc: 0.8650\n",
      "Epoch 314/3000\n",
      " - 0s - loss: 0.3269 - mean_squared_error: 0.0888 - acc: 0.9230 - val_loss: 0.3848 - val_mean_squared_error: 0.1139 - val_acc: 0.8650\n",
      "Epoch 315/3000\n",
      " - 0s - loss: 0.3289 - mean_squared_error: 0.0898 - acc: 0.9218 - val_loss: 0.3844 - val_mean_squared_error: 0.1138 - val_acc: 0.8650\n",
      "Epoch 316/3000\n",
      " - 0s - loss: 0.3242 - mean_squared_error: 0.0877 - acc: 0.9279 - val_loss: 0.3840 - val_mean_squared_error: 0.1136 - val_acc: 0.8650\n",
      "Epoch 317/3000\n",
      " - 0s - loss: 0.3265 - mean_squared_error: 0.0889 - acc: 0.9218 - val_loss: 0.3838 - val_mean_squared_error: 0.1136 - val_acc: 0.8650\n",
      "Epoch 318/3000\n",
      " - 0s - loss: 0.3253 - mean_squared_error: 0.0883 - acc: 0.9230 - val_loss: 0.3835 - val_mean_squared_error: 0.1135 - val_acc: 0.8650\n",
      "Epoch 319/3000\n",
      " - 0s - loss: 0.3172 - mean_squared_error: 0.0854 - acc: 0.9218 - val_loss: 0.3834 - val_mean_squared_error: 0.1134 - val_acc: 0.8650\n",
      "Epoch 320/3000\n",
      " - 0s - loss: 0.3230 - mean_squared_error: 0.0869 - acc: 0.9242 - val_loss: 0.3833 - val_mean_squared_error: 0.1134 - val_acc: 0.8650\n",
      "Epoch 321/3000\n",
      " - 0s - loss: 0.3245 - mean_squared_error: 0.0880 - acc: 0.9242 - val_loss: 0.3832 - val_mean_squared_error: 0.1134 - val_acc: 0.8650\n",
      "Epoch 322/3000\n",
      " - 0s - loss: 0.3245 - mean_squared_error: 0.0882 - acc: 0.9254 - val_loss: 0.3831 - val_mean_squared_error: 0.1133 - val_acc: 0.8650\n",
      "Epoch 323/3000\n",
      " - 0s - loss: 0.3206 - mean_squared_error: 0.0864 - acc: 0.9291 - val_loss: 0.3829 - val_mean_squared_error: 0.1133 - val_acc: 0.8650\n",
      "Epoch 324/3000\n",
      " - 0s - loss: 0.3224 - mean_squared_error: 0.0874 - acc: 0.9230 - val_loss: 0.3826 - val_mean_squared_error: 0.1132 - val_acc: 0.8650\n",
      "Epoch 325/3000\n",
      " - 0s - loss: 0.3215 - mean_squared_error: 0.0866 - acc: 0.9169 - val_loss: 0.3823 - val_mean_squared_error: 0.1131 - val_acc: 0.8650\n",
      "Epoch 326/3000\n",
      " - 0s - loss: 0.3219 - mean_squared_error: 0.0869 - acc: 0.9230 - val_loss: 0.3820 - val_mean_squared_error: 0.1130 - val_acc: 0.8650\n",
      "Epoch 327/3000\n",
      " - 0s - loss: 0.3202 - mean_squared_error: 0.0864 - acc: 0.9279 - val_loss: 0.3816 - val_mean_squared_error: 0.1128 - val_acc: 0.8650\n",
      "Epoch 328/3000\n",
      " - 0s - loss: 0.3195 - mean_squared_error: 0.0864 - acc: 0.9218 - val_loss: 0.3811 - val_mean_squared_error: 0.1127 - val_acc: 0.8650\n",
      "Epoch 329/3000\n",
      " - 0s - loss: 0.3165 - mean_squared_error: 0.0850 - acc: 0.9291 - val_loss: 0.3806 - val_mean_squared_error: 0.1125 - val_acc: 0.8650\n",
      "Epoch 330/3000\n",
      " - 0s - loss: 0.3199 - mean_squared_error: 0.0865 - acc: 0.9254 - val_loss: 0.3800 - val_mean_squared_error: 0.1123 - val_acc: 0.8650\n",
      "Epoch 331/3000\n",
      " - 0s - loss: 0.3191 - mean_squared_error: 0.0860 - acc: 0.9279 - val_loss: 0.3794 - val_mean_squared_error: 0.1121 - val_acc: 0.8650\n",
      "Epoch 332/3000\n",
      " - 0s - loss: 0.3148 - mean_squared_error: 0.0846 - acc: 0.9230 - val_loss: 0.3789 - val_mean_squared_error: 0.1119 - val_acc: 0.8650\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 333/3000\n",
      " - 0s - loss: 0.3227 - mean_squared_error: 0.0878 - acc: 0.9242 - val_loss: 0.3784 - val_mean_squared_error: 0.1118 - val_acc: 0.8650\n",
      "Epoch 334/3000\n",
      " - 0s - loss: 0.3162 - mean_squared_error: 0.0852 - acc: 0.9181 - val_loss: 0.3780 - val_mean_squared_error: 0.1116 - val_acc: 0.8650\n",
      "Epoch 335/3000\n",
      " - 0s - loss: 0.3163 - mean_squared_error: 0.0852 - acc: 0.9230 - val_loss: 0.3776 - val_mean_squared_error: 0.1115 - val_acc: 0.8650\n",
      "Epoch 336/3000\n",
      " - 0s - loss: 0.3189 - mean_squared_error: 0.0859 - acc: 0.9242 - val_loss: 0.3773 - val_mean_squared_error: 0.1114 - val_acc: 0.8650\n",
      "Epoch 337/3000\n",
      " - 0s - loss: 0.3132 - mean_squared_error: 0.0841 - acc: 0.9242 - val_loss: 0.3771 - val_mean_squared_error: 0.1113 - val_acc: 0.8650\n",
      "Epoch 338/3000\n",
      " - 0s - loss: 0.3109 - mean_squared_error: 0.0829 - acc: 0.9315 - val_loss: 0.3769 - val_mean_squared_error: 0.1112 - val_acc: 0.8650\n",
      "Epoch 339/3000\n",
      " - 0s - loss: 0.3103 - mean_squared_error: 0.0830 - acc: 0.9230 - val_loss: 0.3767 - val_mean_squared_error: 0.1112 - val_acc: 0.8650\n",
      "Epoch 340/3000\n",
      " - 0s - loss: 0.3064 - mean_squared_error: 0.0814 - acc: 0.9254 - val_loss: 0.3766 - val_mean_squared_error: 0.1112 - val_acc: 0.8650\n",
      "Epoch 341/3000\n",
      " - 0s - loss: 0.3110 - mean_squared_error: 0.0831 - acc: 0.9303 - val_loss: 0.3766 - val_mean_squared_error: 0.1111 - val_acc: 0.8650\n",
      "Epoch 342/3000\n",
      " - 0s - loss: 0.3112 - mean_squared_error: 0.0831 - acc: 0.9254 - val_loss: 0.3764 - val_mean_squared_error: 0.1111 - val_acc: 0.8650\n",
      "Epoch 343/3000\n",
      " - 0s - loss: 0.3094 - mean_squared_error: 0.0825 - acc: 0.9193 - val_loss: 0.3763 - val_mean_squared_error: 0.1110 - val_acc: 0.8650\n",
      "Epoch 344/3000\n",
      " - 0s - loss: 0.3112 - mean_squared_error: 0.0831 - acc: 0.9291 - val_loss: 0.3760 - val_mean_squared_error: 0.1110 - val_acc: 0.8650\n",
      "Epoch 345/3000\n",
      " - 0s - loss: 0.3100 - mean_squared_error: 0.0830 - acc: 0.9242 - val_loss: 0.3758 - val_mean_squared_error: 0.1109 - val_acc: 0.8650\n",
      "Epoch 346/3000\n",
      " - 0s - loss: 0.3121 - mean_squared_error: 0.0837 - acc: 0.9181 - val_loss: 0.3754 - val_mean_squared_error: 0.1108 - val_acc: 0.8650\n",
      "Epoch 347/3000\n",
      " - 0s - loss: 0.3116 - mean_squared_error: 0.0840 - acc: 0.9279 - val_loss: 0.3751 - val_mean_squared_error: 0.1106 - val_acc: 0.8650\n",
      "Epoch 348/3000\n",
      " - 0s - loss: 0.3090 - mean_squared_error: 0.0830 - acc: 0.9144 - val_loss: 0.3746 - val_mean_squared_error: 0.1105 - val_acc: 0.8650\n",
      "Epoch 349/3000\n",
      " - 0s - loss: 0.3105 - mean_squared_error: 0.0830 - acc: 0.9279 - val_loss: 0.3742 - val_mean_squared_error: 0.1104 - val_acc: 0.8650\n",
      "Epoch 350/3000\n",
      " - 0s - loss: 0.3108 - mean_squared_error: 0.0832 - acc: 0.9230 - val_loss: 0.3736 - val_mean_squared_error: 0.1102 - val_acc: 0.8650\n",
      "Epoch 351/3000\n",
      " - 0s - loss: 0.3099 - mean_squared_error: 0.0834 - acc: 0.9230 - val_loss: 0.3731 - val_mean_squared_error: 0.1100 - val_acc: 0.8650\n",
      "Epoch 352/3000\n",
      " - 0s - loss: 0.3060 - mean_squared_error: 0.0814 - acc: 0.9267 - val_loss: 0.3726 - val_mean_squared_error: 0.1098 - val_acc: 0.8650\n",
      "Epoch 353/3000\n",
      " - 0s - loss: 0.3044 - mean_squared_error: 0.0811 - acc: 0.9242 - val_loss: 0.3722 - val_mean_squared_error: 0.1097 - val_acc: 0.8650\n",
      "Epoch 354/3000\n",
      " - 0s - loss: 0.3020 - mean_squared_error: 0.0803 - acc: 0.9267 - val_loss: 0.3718 - val_mean_squared_error: 0.1096 - val_acc: 0.8650\n",
      "Epoch 355/3000\n",
      " - 0s - loss: 0.3067 - mean_squared_error: 0.0820 - acc: 0.9267 - val_loss: 0.3714 - val_mean_squared_error: 0.1095 - val_acc: 0.8650\n",
      "Epoch 356/3000\n",
      " - 0s - loss: 0.3089 - mean_squared_error: 0.0829 - acc: 0.9205 - val_loss: 0.3712 - val_mean_squared_error: 0.1094 - val_acc: 0.8650\n",
      "Epoch 357/3000\n",
      " - 0s - loss: 0.3051 - mean_squared_error: 0.0815 - acc: 0.9291 - val_loss: 0.3709 - val_mean_squared_error: 0.1093 - val_acc: 0.8650\n",
      "Epoch 358/3000\n",
      " - 0s - loss: 0.3059 - mean_squared_error: 0.0818 - acc: 0.9181 - val_loss: 0.3707 - val_mean_squared_error: 0.1092 - val_acc: 0.8650\n",
      "Epoch 359/3000\n",
      " - 0s - loss: 0.3000 - mean_squared_error: 0.0788 - acc: 0.9279 - val_loss: 0.3706 - val_mean_squared_error: 0.1092 - val_acc: 0.8650\n",
      "Epoch 360/3000\n",
      " - 0s - loss: 0.2994 - mean_squared_error: 0.0792 - acc: 0.9279 - val_loss: 0.3705 - val_mean_squared_error: 0.1092 - val_acc: 0.8650\n",
      "Epoch 361/3000\n",
      " - 0s - loss: 0.3090 - mean_squared_error: 0.0829 - acc: 0.9279 - val_loss: 0.3705 - val_mean_squared_error: 0.1091 - val_acc: 0.8650\n",
      "Epoch 362/3000\n",
      " - 0s - loss: 0.3070 - mean_squared_error: 0.0819 - acc: 0.9254 - val_loss: 0.3703 - val_mean_squared_error: 0.1091 - val_acc: 0.8650\n",
      "Epoch 363/3000\n",
      " - 0s - loss: 0.3061 - mean_squared_error: 0.0815 - acc: 0.9303 - val_loss: 0.3702 - val_mean_squared_error: 0.1091 - val_acc: 0.8650\n",
      "Epoch 364/3000\n",
      " - 0s - loss: 0.3002 - mean_squared_error: 0.0796 - acc: 0.9315 - val_loss: 0.3700 - val_mean_squared_error: 0.1090 - val_acc: 0.8650\n",
      "Epoch 365/3000\n",
      " - 0s - loss: 0.3006 - mean_squared_error: 0.0795 - acc: 0.9218 - val_loss: 0.3697 - val_mean_squared_error: 0.1089 - val_acc: 0.8650\n",
      "Epoch 366/3000\n",
      " - 0s - loss: 0.3049 - mean_squared_error: 0.0811 - acc: 0.9291 - val_loss: 0.3694 - val_mean_squared_error: 0.1088 - val_acc: 0.8650\n",
      "Epoch 367/3000\n",
      " - 0s - loss: 0.3019 - mean_squared_error: 0.0800 - acc: 0.9254 - val_loss: 0.3691 - val_mean_squared_error: 0.1087 - val_acc: 0.8650\n",
      "Epoch 368/3000\n",
      " - 0s - loss: 0.3009 - mean_squared_error: 0.0794 - acc: 0.9352 - val_loss: 0.3687 - val_mean_squared_error: 0.1086 - val_acc: 0.8650\n",
      "Epoch 369/3000\n",
      " - 0s - loss: 0.2993 - mean_squared_error: 0.0792 - acc: 0.9315 - val_loss: 0.3682 - val_mean_squared_error: 0.1084 - val_acc: 0.8650\n",
      "Epoch 370/3000\n",
      " - 0s - loss: 0.3012 - mean_squared_error: 0.0799 - acc: 0.9291 - val_loss: 0.3678 - val_mean_squared_error: 0.1083 - val_acc: 0.8650\n",
      "Epoch 371/3000\n",
      " - 0s - loss: 0.2990 - mean_squared_error: 0.0793 - acc: 0.9267 - val_loss: 0.3673 - val_mean_squared_error: 0.1081 - val_acc: 0.8650\n",
      "Epoch 372/3000\n",
      " - 0s - loss: 0.3005 - mean_squared_error: 0.0798 - acc: 0.9291 - val_loss: 0.3668 - val_mean_squared_error: 0.1080 - val_acc: 0.8650\n",
      "Epoch 373/3000\n",
      " - 0s - loss: 0.2951 - mean_squared_error: 0.0775 - acc: 0.9230 - val_loss: 0.3664 - val_mean_squared_error: 0.1078 - val_acc: 0.8650\n",
      "Epoch 374/3000\n",
      " - 0s - loss: 0.2972 - mean_squared_error: 0.0790 - acc: 0.9267 - val_loss: 0.3660 - val_mean_squared_error: 0.1077 - val_acc: 0.8650\n",
      "Epoch 375/3000\n",
      " - 0s - loss: 0.2963 - mean_squared_error: 0.0785 - acc: 0.9254 - val_loss: 0.3657 - val_mean_squared_error: 0.1076 - val_acc: 0.8650\n",
      "Epoch 376/3000\n",
      " - 0s - loss: 0.2959 - mean_squared_error: 0.0780 - acc: 0.9254 - val_loss: 0.3655 - val_mean_squared_error: 0.1075 - val_acc: 0.8650\n",
      "Epoch 377/3000\n",
      " - 0s - loss: 0.2895 - mean_squared_error: 0.0757 - acc: 0.9242 - val_loss: 0.3653 - val_mean_squared_error: 0.1075 - val_acc: 0.8650\n",
      "Epoch 378/3000\n",
      " - 0s - loss: 0.2878 - mean_squared_error: 0.0752 - acc: 0.9328 - val_loss: 0.3651 - val_mean_squared_error: 0.1074 - val_acc: 0.8650\n",
      "Epoch 379/3000\n",
      " - 0s - loss: 0.2934 - mean_squared_error: 0.0770 - acc: 0.9267 - val_loss: 0.3650 - val_mean_squared_error: 0.1074 - val_acc: 0.8650\n",
      "Epoch 380/3000\n",
      " - 0s - loss: 0.2944 - mean_squared_error: 0.0776 - acc: 0.9254 - val_loss: 0.3649 - val_mean_squared_error: 0.1074 - val_acc: 0.8650\n",
      "Epoch 381/3000\n",
      " - 0s - loss: 0.2938 - mean_squared_error: 0.0769 - acc: 0.9279 - val_loss: 0.3648 - val_mean_squared_error: 0.1073 - val_acc: 0.8650\n",
      "Epoch 382/3000\n",
      " - 0s - loss: 0.2903 - mean_squared_error: 0.0761 - acc: 0.9254 - val_loss: 0.3647 - val_mean_squared_error: 0.1073 - val_acc: 0.8650\n",
      "Epoch 383/3000\n",
      " - 0s - loss: 0.2952 - mean_squared_error: 0.0779 - acc: 0.9279 - val_loss: 0.3646 - val_mean_squared_error: 0.1073 - val_acc: 0.8650\n",
      "Epoch 384/3000\n",
      " - 0s - loss: 0.2985 - mean_squared_error: 0.0795 - acc: 0.9169 - val_loss: 0.3644 - val_mean_squared_error: 0.1072 - val_acc: 0.8650\n",
      "Epoch 385/3000\n",
      " - 0s - loss: 0.2914 - mean_squared_error: 0.0767 - acc: 0.9352 - val_loss: 0.3642 - val_mean_squared_error: 0.1071 - val_acc: 0.8650\n",
      "Epoch 386/3000\n",
      " - 0s - loss: 0.2937 - mean_squared_error: 0.0776 - acc: 0.9291 - val_loss: 0.3639 - val_mean_squared_error: 0.1070 - val_acc: 0.8650\n",
      "Epoch 387/3000\n",
      " - 0s - loss: 0.2958 - mean_squared_error: 0.0784 - acc: 0.9242 - val_loss: 0.3636 - val_mean_squared_error: 0.1069 - val_acc: 0.8650\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 388/3000\n",
      " - 0s - loss: 0.2925 - mean_squared_error: 0.0771 - acc: 0.9303 - val_loss: 0.3632 - val_mean_squared_error: 0.1068 - val_acc: 0.8650\n",
      "Epoch 389/3000\n",
      " - 0s - loss: 0.2936 - mean_squared_error: 0.0774 - acc: 0.9315 - val_loss: 0.3628 - val_mean_squared_error: 0.1067 - val_acc: 0.8650\n",
      "Epoch 390/3000\n",
      " - 0s - loss: 0.2960 - mean_squared_error: 0.0786 - acc: 0.9303 - val_loss: 0.3624 - val_mean_squared_error: 0.1066 - val_acc: 0.8650\n",
      "Epoch 391/3000\n",
      " - 0s - loss: 0.2864 - mean_squared_error: 0.0746 - acc: 0.9401 - val_loss: 0.3619 - val_mean_squared_error: 0.1064 - val_acc: 0.8650\n",
      "Epoch 392/3000\n",
      " - 0s - loss: 0.2902 - mean_squared_error: 0.0759 - acc: 0.9328 - val_loss: 0.3615 - val_mean_squared_error: 0.1063 - val_acc: 0.8650\n",
      "Epoch 393/3000\n",
      " - 0s - loss: 0.2921 - mean_squared_error: 0.0773 - acc: 0.9218 - val_loss: 0.3611 - val_mean_squared_error: 0.1062 - val_acc: 0.8650\n",
      "Epoch 394/3000\n",
      " - 0s - loss: 0.2890 - mean_squared_error: 0.0758 - acc: 0.9279 - val_loss: 0.3608 - val_mean_squared_error: 0.1061 - val_acc: 0.8650\n",
      "Epoch 395/3000\n",
      " - 0s - loss: 0.2877 - mean_squared_error: 0.0752 - acc: 0.9389 - val_loss: 0.3605 - val_mean_squared_error: 0.1060 - val_acc: 0.8650\n",
      "Epoch 396/3000\n",
      " - 0s - loss: 0.2837 - mean_squared_error: 0.0734 - acc: 0.9377 - val_loss: 0.3603 - val_mean_squared_error: 0.1059 - val_acc: 0.8650\n",
      "Epoch 397/3000\n",
      " - 0s - loss: 0.2898 - mean_squared_error: 0.0762 - acc: 0.9279 - val_loss: 0.3601 - val_mean_squared_error: 0.1059 - val_acc: 0.8650\n",
      "Epoch 398/3000\n",
      " - 0s - loss: 0.2825 - mean_squared_error: 0.0732 - acc: 0.9352 - val_loss: 0.3599 - val_mean_squared_error: 0.1058 - val_acc: 0.8650\n",
      "Epoch 399/3000\n",
      " - 0s - loss: 0.2852 - mean_squared_error: 0.0744 - acc: 0.9328 - val_loss: 0.3598 - val_mean_squared_error: 0.1058 - val_acc: 0.8650\n",
      "Epoch 400/3000\n",
      " - 0s - loss: 0.2839 - mean_squared_error: 0.0738 - acc: 0.9328 - val_loss: 0.3597 - val_mean_squared_error: 0.1058 - val_acc: 0.8650\n",
      "Epoch 401/3000\n",
      " - 0s - loss: 0.2898 - mean_squared_error: 0.0762 - acc: 0.9340 - val_loss: 0.3597 - val_mean_squared_error: 0.1057 - val_acc: 0.8650\n",
      "Epoch 402/3000\n",
      " - 0s - loss: 0.2849 - mean_squared_error: 0.0745 - acc: 0.9291 - val_loss: 0.3596 - val_mean_squared_error: 0.1057 - val_acc: 0.8650\n",
      "Epoch 403/3000\n",
      " - 0s - loss: 0.2891 - mean_squared_error: 0.0763 - acc: 0.9267 - val_loss: 0.3595 - val_mean_squared_error: 0.1057 - val_acc: 0.8650\n",
      "Epoch 404/3000\n",
      " - 0s - loss: 0.2808 - mean_squared_error: 0.0724 - acc: 0.9340 - val_loss: 0.3593 - val_mean_squared_error: 0.1056 - val_acc: 0.8650\n",
      "Epoch 405/3000\n",
      " - 0s - loss: 0.2854 - mean_squared_error: 0.0748 - acc: 0.9242 - val_loss: 0.3591 - val_mean_squared_error: 0.1055 - val_acc: 0.8650\n",
      "Epoch 406/3000\n",
      " - 0s - loss: 0.2865 - mean_squared_error: 0.0751 - acc: 0.9267 - val_loss: 0.3588 - val_mean_squared_error: 0.1055 - val_acc: 0.8650\n",
      "Epoch 407/3000\n",
      " - 0s - loss: 0.2875 - mean_squared_error: 0.0758 - acc: 0.9218 - val_loss: 0.3585 - val_mean_squared_error: 0.1054 - val_acc: 0.8650\n",
      "Epoch 408/3000\n",
      " - 0s - loss: 0.2823 - mean_squared_error: 0.0734 - acc: 0.9279 - val_loss: 0.3582 - val_mean_squared_error: 0.1053 - val_acc: 0.8650\n",
      "Epoch 409/3000\n",
      " - 0s - loss: 0.2826 - mean_squared_error: 0.0739 - acc: 0.9279 - val_loss: 0.3578 - val_mean_squared_error: 0.1052 - val_acc: 0.8650\n",
      "Epoch 410/3000\n",
      " - 0s - loss: 0.2843 - mean_squared_error: 0.0745 - acc: 0.9279 - val_loss: 0.3574 - val_mean_squared_error: 0.1050 - val_acc: 0.8650\n",
      "Epoch 411/3000\n",
      " - 0s - loss: 0.2819 - mean_squared_error: 0.0734 - acc: 0.9340 - val_loss: 0.3570 - val_mean_squared_error: 0.1049 - val_acc: 0.8650\n",
      "Epoch 412/3000\n",
      " - 0s - loss: 0.2806 - mean_squared_error: 0.0728 - acc: 0.9328 - val_loss: 0.3566 - val_mean_squared_error: 0.1048 - val_acc: 0.8650\n",
      "Epoch 413/3000\n",
      " - 0s - loss: 0.2777 - mean_squared_error: 0.0713 - acc: 0.9487 - val_loss: 0.3563 - val_mean_squared_error: 0.1047 - val_acc: 0.8650\n",
      "Epoch 414/3000\n",
      " - 0s - loss: 0.2820 - mean_squared_error: 0.0732 - acc: 0.9352 - val_loss: 0.3560 - val_mean_squared_error: 0.1046 - val_acc: 0.8650\n",
      "Epoch 415/3000\n",
      " - 0s - loss: 0.2796 - mean_squared_error: 0.0722 - acc: 0.9315 - val_loss: 0.3557 - val_mean_squared_error: 0.1045 - val_acc: 0.8650\n",
      "Epoch 416/3000\n",
      " - 0s - loss: 0.2818 - mean_squared_error: 0.0739 - acc: 0.9315 - val_loss: 0.3555 - val_mean_squared_error: 0.1045 - val_acc: 0.8650\n",
      "Epoch 417/3000\n",
      " - 0s - loss: 0.2775 - mean_squared_error: 0.0714 - acc: 0.9315 - val_loss: 0.3553 - val_mean_squared_error: 0.1044 - val_acc: 0.8650\n",
      "Epoch 418/3000\n",
      " - 0s - loss: 0.2796 - mean_squared_error: 0.0730 - acc: 0.9254 - val_loss: 0.3552 - val_mean_squared_error: 0.1044 - val_acc: 0.8650\n",
      "Epoch 419/3000\n",
      " - 0s - loss: 0.2752 - mean_squared_error: 0.0713 - acc: 0.9303 - val_loss: 0.3551 - val_mean_squared_error: 0.1043 - val_acc: 0.8650\n",
      "Epoch 420/3000\n",
      " - 0s - loss: 0.2777 - mean_squared_error: 0.0724 - acc: 0.9303 - val_loss: 0.3550 - val_mean_squared_error: 0.1043 - val_acc: 0.8650\n",
      "Epoch 421/3000\n",
      " - 0s - loss: 0.2775 - mean_squared_error: 0.0716 - acc: 0.9303 - val_loss: 0.3550 - val_mean_squared_error: 0.1043 - val_acc: 0.8650\n",
      "Epoch 422/3000\n",
      " - 0s - loss: 0.2712 - mean_squared_error: 0.0695 - acc: 0.9377 - val_loss: 0.3549 - val_mean_squared_error: 0.1043 - val_acc: 0.8650\n",
      "Epoch 423/3000\n",
      " - 0s - loss: 0.2800 - mean_squared_error: 0.0738 - acc: 0.9242 - val_loss: 0.3547 - val_mean_squared_error: 0.1042 - val_acc: 0.8650\n",
      "Epoch 424/3000\n",
      " - 0s - loss: 0.2748 - mean_squared_error: 0.0714 - acc: 0.9401 - val_loss: 0.3546 - val_mean_squared_error: 0.1042 - val_acc: 0.8650\n",
      "Epoch 425/3000\n",
      " - 0s - loss: 0.2793 - mean_squared_error: 0.0731 - acc: 0.9279 - val_loss: 0.3544 - val_mean_squared_error: 0.1041 - val_acc: 0.8650\n",
      "Epoch 426/3000\n",
      " - 0s - loss: 0.2776 - mean_squared_error: 0.0719 - acc: 0.9328 - val_loss: 0.3542 - val_mean_squared_error: 0.1041 - val_acc: 0.8650\n",
      "Epoch 427/3000\n",
      " - 0s - loss: 0.2740 - mean_squared_error: 0.0712 - acc: 0.9340 - val_loss: 0.3539 - val_mean_squared_error: 0.1040 - val_acc: 0.8650\n",
      "Epoch 428/3000\n",
      " - 0s - loss: 0.2810 - mean_squared_error: 0.0733 - acc: 0.9254 - val_loss: 0.3536 - val_mean_squared_error: 0.1039 - val_acc: 0.8650\n",
      "Epoch 429/3000\n",
      " - 0s - loss: 0.2764 - mean_squared_error: 0.0719 - acc: 0.9303 - val_loss: 0.3533 - val_mean_squared_error: 0.1038 - val_acc: 0.8650\n",
      "Epoch 430/3000\n",
      " - 0s - loss: 0.2757 - mean_squared_error: 0.0713 - acc: 0.9254 - val_loss: 0.3529 - val_mean_squared_error: 0.1037 - val_acc: 0.8650\n",
      "Epoch 431/3000\n",
      " - 0s - loss: 0.2741 - mean_squared_error: 0.0707 - acc: 0.9315 - val_loss: 0.3525 - val_mean_squared_error: 0.1036 - val_acc: 0.8650\n",
      "Epoch 432/3000\n",
      " - 0s - loss: 0.2762 - mean_squared_error: 0.0718 - acc: 0.9267 - val_loss: 0.3522 - val_mean_squared_error: 0.1035 - val_acc: 0.8650\n",
      "Epoch 433/3000\n",
      " - 0s - loss: 0.2708 - mean_squared_error: 0.0698 - acc: 0.9340 - val_loss: 0.3519 - val_mean_squared_error: 0.1034 - val_acc: 0.8650\n",
      "Epoch 434/3000\n",
      " - 0s - loss: 0.2730 - mean_squared_error: 0.0703 - acc: 0.9340 - val_loss: 0.3516 - val_mean_squared_error: 0.1033 - val_acc: 0.8650\n",
      "Epoch 435/3000\n",
      " - 0s - loss: 0.2750 - mean_squared_error: 0.0714 - acc: 0.9340 - val_loss: 0.3513 - val_mean_squared_error: 0.1032 - val_acc: 0.8650\n",
      "Epoch 436/3000\n",
      " - 0s - loss: 0.2724 - mean_squared_error: 0.0702 - acc: 0.9328 - val_loss: 0.3511 - val_mean_squared_error: 0.1032 - val_acc: 0.8650\n",
      "Epoch 437/3000\n",
      " - 0s - loss: 0.2751 - mean_squared_error: 0.0716 - acc: 0.9303 - val_loss: 0.3510 - val_mean_squared_error: 0.1031 - val_acc: 0.8650\n",
      "Epoch 438/3000\n",
      " - 0s - loss: 0.2707 - mean_squared_error: 0.0698 - acc: 0.9328 - val_loss: 0.3508 - val_mean_squared_error: 0.1031 - val_acc: 0.8650\n",
      "Epoch 439/3000\n",
      " - 0s - loss: 0.2667 - mean_squared_error: 0.0684 - acc: 0.9328 - val_loss: 0.3507 - val_mean_squared_error: 0.1030 - val_acc: 0.8650\n",
      "Epoch 440/3000\n",
      " - 0s - loss: 0.2705 - mean_squared_error: 0.0697 - acc: 0.9340 - val_loss: 0.3507 - val_mean_squared_error: 0.1030 - val_acc: 0.8650\n",
      "Epoch 441/3000\n",
      " - 0s - loss: 0.2670 - mean_squared_error: 0.0683 - acc: 0.9340 - val_loss: 0.3506 - val_mean_squared_error: 0.1030 - val_acc: 0.8650\n",
      "Epoch 442/3000\n",
      " - 0s - loss: 0.2679 - mean_squared_error: 0.0685 - acc: 0.9389 - val_loss: 0.3506 - val_mean_squared_error: 0.1030 - val_acc: 0.8650\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 443/3000\n",
      " - 0s - loss: 0.2709 - mean_squared_error: 0.0694 - acc: 0.9438 - val_loss: 0.3504 - val_mean_squared_error: 0.1030 - val_acc: 0.8650\n",
      "Epoch 444/3000\n",
      " - 0s - loss: 0.2686 - mean_squared_error: 0.0691 - acc: 0.9340 - val_loss: 0.3503 - val_mean_squared_error: 0.1029 - val_acc: 0.8650\n",
      "Epoch 445/3000\n",
      " - 0s - loss: 0.2701 - mean_squared_error: 0.0697 - acc: 0.9364 - val_loss: 0.3501 - val_mean_squared_error: 0.1029 - val_acc: 0.8650\n",
      "Epoch 446/3000\n",
      " - 0s - loss: 0.2640 - mean_squared_error: 0.0670 - acc: 0.9413 - val_loss: 0.3499 - val_mean_squared_error: 0.1028 - val_acc: 0.8650\n",
      "Epoch 447/3000\n",
      " - 0s - loss: 0.2690 - mean_squared_error: 0.0692 - acc: 0.9389 - val_loss: 0.3496 - val_mean_squared_error: 0.1027 - val_acc: 0.8650\n",
      "Epoch 448/3000\n",
      " - 0s - loss: 0.2676 - mean_squared_error: 0.0685 - acc: 0.9364 - val_loss: 0.3494 - val_mean_squared_error: 0.1026 - val_acc: 0.8650\n",
      "Epoch 449/3000\n",
      " - 0s - loss: 0.2666 - mean_squared_error: 0.0684 - acc: 0.9303 - val_loss: 0.3490 - val_mean_squared_error: 0.1025 - val_acc: 0.8650\n",
      "Epoch 450/3000\n",
      " - 0s - loss: 0.2682 - mean_squared_error: 0.0688 - acc: 0.9377 - val_loss: 0.3487 - val_mean_squared_error: 0.1024 - val_acc: 0.8650\n",
      "Epoch 451/3000\n",
      " - 0s - loss: 0.2624 - mean_squared_error: 0.0668 - acc: 0.9340 - val_loss: 0.3483 - val_mean_squared_error: 0.1023 - val_acc: 0.8650\n",
      "Epoch 452/3000\n",
      " - 0s - loss: 0.2674 - mean_squared_error: 0.0687 - acc: 0.9340 - val_loss: 0.3480 - val_mean_squared_error: 0.1022 - val_acc: 0.8650\n",
      "Epoch 453/3000\n",
      " - 0s - loss: 0.2581 - mean_squared_error: 0.0653 - acc: 0.9377 - val_loss: 0.3477 - val_mean_squared_error: 0.1022 - val_acc: 0.8650\n",
      "Epoch 454/3000\n",
      " - 0s - loss: 0.2674 - mean_squared_error: 0.0689 - acc: 0.9279 - val_loss: 0.3475 - val_mean_squared_error: 0.1021 - val_acc: 0.8650\n",
      "Epoch 455/3000\n",
      " - 0s - loss: 0.2664 - mean_squared_error: 0.0681 - acc: 0.9425 - val_loss: 0.3472 - val_mean_squared_error: 0.1020 - val_acc: 0.8650\n",
      "Epoch 456/3000\n",
      " - 0s - loss: 0.2639 - mean_squared_error: 0.0673 - acc: 0.9315 - val_loss: 0.3470 - val_mean_squared_error: 0.1020 - val_acc: 0.8650\n",
      "Epoch 457/3000\n",
      " - 0s - loss: 0.2629 - mean_squared_error: 0.0673 - acc: 0.9364 - val_loss: 0.3469 - val_mean_squared_error: 0.1019 - val_acc: 0.8650\n",
      "Epoch 458/3000\n",
      " - 0s - loss: 0.2648 - mean_squared_error: 0.0679 - acc: 0.9364 - val_loss: 0.3468 - val_mean_squared_error: 0.1019 - val_acc: 0.8650\n",
      "Epoch 459/3000\n",
      " - 0s - loss: 0.2639 - mean_squared_error: 0.0681 - acc: 0.9352 - val_loss: 0.3467 - val_mean_squared_error: 0.1019 - val_acc: 0.8650\n",
      "Epoch 460/3000\n",
      " - 0s - loss: 0.2590 - mean_squared_error: 0.0655 - acc: 0.9377 - val_loss: 0.3466 - val_mean_squared_error: 0.1018 - val_acc: 0.8650\n",
      "Epoch 461/3000\n",
      " - 0s - loss: 0.2666 - mean_squared_error: 0.0688 - acc: 0.9340 - val_loss: 0.3466 - val_mean_squared_error: 0.1018 - val_acc: 0.8650\n",
      "Epoch 462/3000\n",
      " - 0s - loss: 0.2671 - mean_squared_error: 0.0691 - acc: 0.9267 - val_loss: 0.3465 - val_mean_squared_error: 0.1018 - val_acc: 0.8650\n",
      "Epoch 463/3000\n",
      " - 0s - loss: 0.2642 - mean_squared_error: 0.0676 - acc: 0.9315 - val_loss: 0.3464 - val_mean_squared_error: 0.1018 - val_acc: 0.8650\n",
      "Epoch 464/3000\n",
      " - 0s - loss: 0.2618 - mean_squared_error: 0.0667 - acc: 0.9389 - val_loss: 0.3463 - val_mean_squared_error: 0.1017 - val_acc: 0.8650\n",
      "Epoch 465/3000\n",
      " - 0s - loss: 0.2614 - mean_squared_error: 0.0665 - acc: 0.9340 - val_loss: 0.3461 - val_mean_squared_error: 0.1017 - val_acc: 0.8650\n",
      "Epoch 466/3000\n",
      " - 0s - loss: 0.2605 - mean_squared_error: 0.0665 - acc: 0.9315 - val_loss: 0.3459 - val_mean_squared_error: 0.1016 - val_acc: 0.8650\n",
      "Epoch 467/3000\n",
      " - 0s - loss: 0.2574 - mean_squared_error: 0.0658 - acc: 0.9315 - val_loss: 0.3457 - val_mean_squared_error: 0.1016 - val_acc: 0.8650\n",
      "Epoch 468/3000\n",
      " - 0s - loss: 0.2629 - mean_squared_error: 0.0670 - acc: 0.9377 - val_loss: 0.3454 - val_mean_squared_error: 0.1015 - val_acc: 0.8650\n",
      "Epoch 469/3000\n",
      " - 0s - loss: 0.2618 - mean_squared_error: 0.0670 - acc: 0.9352 - val_loss: 0.3451 - val_mean_squared_error: 0.1014 - val_acc: 0.8650\n",
      "Epoch 470/3000\n",
      " - 0s - loss: 0.2577 - mean_squared_error: 0.0654 - acc: 0.9364 - val_loss: 0.3448 - val_mean_squared_error: 0.1013 - val_acc: 0.8650\n",
      "Epoch 471/3000\n",
      " - 0s - loss: 0.2593 - mean_squared_error: 0.0663 - acc: 0.9315 - val_loss: 0.3445 - val_mean_squared_error: 0.1013 - val_acc: 0.8650\n",
      "Epoch 472/3000\n",
      " - 0s - loss: 0.2575 - mean_squared_error: 0.0657 - acc: 0.9340 - val_loss: 0.3442 - val_mean_squared_error: 0.1012 - val_acc: 0.8650\n",
      "Epoch 473/3000\n",
      " - 0s - loss: 0.2562 - mean_squared_error: 0.0647 - acc: 0.9425 - val_loss: 0.3439 - val_mean_squared_error: 0.1011 - val_acc: 0.8650\n",
      "Epoch 474/3000\n",
      " - 0s - loss: 0.2591 - mean_squared_error: 0.0661 - acc: 0.9352 - val_loss: 0.3437 - val_mean_squared_error: 0.1010 - val_acc: 0.8650\n",
      "Epoch 475/3000\n",
      " - 0s - loss: 0.2579 - mean_squared_error: 0.0655 - acc: 0.9438 - val_loss: 0.3435 - val_mean_squared_error: 0.1010 - val_acc: 0.8650\n",
      "Epoch 476/3000\n",
      " - 0s - loss: 0.2549 - mean_squared_error: 0.0645 - acc: 0.9303 - val_loss: 0.3433 - val_mean_squared_error: 0.1009 - val_acc: 0.8650\n",
      "Epoch 477/3000\n",
      " - 0s - loss: 0.2568 - mean_squared_error: 0.0653 - acc: 0.9352 - val_loss: 0.3432 - val_mean_squared_error: 0.1009 - val_acc: 0.8650\n",
      "Epoch 478/3000\n",
      " - 0s - loss: 0.2565 - mean_squared_error: 0.0650 - acc: 0.9328 - val_loss: 0.3430 - val_mean_squared_error: 0.1009 - val_acc: 0.8650\n",
      "Epoch 479/3000\n",
      " - 0s - loss: 0.2573 - mean_squared_error: 0.0657 - acc: 0.9340 - val_loss: 0.3430 - val_mean_squared_error: 0.1008 - val_acc: 0.8650\n",
      "Epoch 480/3000\n",
      " - 0s - loss: 0.2578 - mean_squared_error: 0.0655 - acc: 0.9389 - val_loss: 0.3429 - val_mean_squared_error: 0.1008 - val_acc: 0.8650\n",
      "Epoch 481/3000\n",
      " - 0s - loss: 0.2543 - mean_squared_error: 0.0646 - acc: 0.9389 - val_loss: 0.3429 - val_mean_squared_error: 0.1008 - val_acc: 0.8650\n",
      "Epoch 482/3000\n",
      " - 0s - loss: 0.2514 - mean_squared_error: 0.0632 - acc: 0.9413 - val_loss: 0.3428 - val_mean_squared_error: 0.1008 - val_acc: 0.8650\n",
      "Epoch 483/3000\n",
      " - 0s - loss: 0.2561 - mean_squared_error: 0.0653 - acc: 0.9328 - val_loss: 0.3427 - val_mean_squared_error: 0.1008 - val_acc: 0.8650\n",
      "Epoch 484/3000\n",
      " - 0s - loss: 0.2568 - mean_squared_error: 0.0655 - acc: 0.9315 - val_loss: 0.3426 - val_mean_squared_error: 0.1007 - val_acc: 0.8650\n",
      "Epoch 485/3000\n",
      " - 0s - loss: 0.2510 - mean_squared_error: 0.0630 - acc: 0.9352 - val_loss: 0.3424 - val_mean_squared_error: 0.1007 - val_acc: 0.8650\n",
      "Epoch 486/3000\n",
      " - 0s - loss: 0.2600 - mean_squared_error: 0.0666 - acc: 0.9291 - val_loss: 0.3422 - val_mean_squared_error: 0.1006 - val_acc: 0.8650\n",
      "Epoch 487/3000\n",
      " - 0s - loss: 0.2495 - mean_squared_error: 0.0630 - acc: 0.9352 - val_loss: 0.3420 - val_mean_squared_error: 0.1006 - val_acc: 0.8650\n",
      "Epoch 488/3000\n",
      " - 0s - loss: 0.2495 - mean_squared_error: 0.0628 - acc: 0.9377 - val_loss: 0.3418 - val_mean_squared_error: 0.1005 - val_acc: 0.8650\n",
      "Epoch 489/3000\n",
      " - 0s - loss: 0.2564 - mean_squared_error: 0.0654 - acc: 0.9425 - val_loss: 0.3415 - val_mean_squared_error: 0.1004 - val_acc: 0.8650\n",
      "Epoch 490/3000\n",
      " - 0s - loss: 0.2559 - mean_squared_error: 0.0650 - acc: 0.9328 - val_loss: 0.3412 - val_mean_squared_error: 0.1004 - val_acc: 0.8650\n",
      "Epoch 491/3000\n",
      " - 0s - loss: 0.2521 - mean_squared_error: 0.0637 - acc: 0.9377 - val_loss: 0.3409 - val_mean_squared_error: 0.1003 - val_acc: 0.8650\n",
      "Epoch 492/3000\n",
      " - 0s - loss: 0.2537 - mean_squared_error: 0.0644 - acc: 0.9413 - val_loss: 0.3407 - val_mean_squared_error: 0.1002 - val_acc: 0.8650\n",
      "Epoch 493/3000\n",
      " - 0s - loss: 0.2520 - mean_squared_error: 0.0642 - acc: 0.9328 - val_loss: 0.3404 - val_mean_squared_error: 0.1001 - val_acc: 0.8650\n",
      "Epoch 494/3000\n",
      " - 0s - loss: 0.2481 - mean_squared_error: 0.0619 - acc: 0.9413 - val_loss: 0.3402 - val_mean_squared_error: 0.1001 - val_acc: 0.8650\n",
      "Epoch 495/3000\n",
      " - 0s - loss: 0.2543 - mean_squared_error: 0.0647 - acc: 0.9425 - val_loss: 0.3400 - val_mean_squared_error: 0.1000 - val_acc: 0.8650\n",
      "Epoch 496/3000\n",
      " - 0s - loss: 0.2506 - mean_squared_error: 0.0639 - acc: 0.9340 - val_loss: 0.3399 - val_mean_squared_error: 0.1000 - val_acc: 0.8650\n",
      "Epoch 497/3000\n",
      " - 0s - loss: 0.2500 - mean_squared_error: 0.0636 - acc: 0.9364 - val_loss: 0.3397 - val_mean_squared_error: 0.0999 - val_acc: 0.8650\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 498/3000\n",
      " - 0s - loss: 0.2498 - mean_squared_error: 0.0634 - acc: 0.9328 - val_loss: 0.3396 - val_mean_squared_error: 0.0999 - val_acc: 0.8650\n",
      "Epoch 499/3000\n",
      " - 0s - loss: 0.2510 - mean_squared_error: 0.0636 - acc: 0.9377 - val_loss: 0.3395 - val_mean_squared_error: 0.0999 - val_acc: 0.8650\n",
      "Epoch 500/3000\n",
      " - 0s - loss: 0.2494 - mean_squared_error: 0.0629 - acc: 0.9389 - val_loss: 0.3395 - val_mean_squared_error: 0.0999 - val_acc: 0.8650\n",
      "Epoch 501/3000\n",
      " - 0s - loss: 0.2488 - mean_squared_error: 0.0628 - acc: 0.9401 - val_loss: 0.3395 - val_mean_squared_error: 0.0999 - val_acc: 0.8650\n",
      "Epoch 502/3000\n",
      " - 0s - loss: 0.2478 - mean_squared_error: 0.0624 - acc: 0.9413 - val_loss: 0.3394 - val_mean_squared_error: 0.0999 - val_acc: 0.8650\n",
      "Epoch 503/3000\n",
      " - 0s - loss: 0.2554 - mean_squared_error: 0.0653 - acc: 0.9364 - val_loss: 0.3393 - val_mean_squared_error: 0.0998 - val_acc: 0.8650\n",
      "Epoch 504/3000\n",
      " - 0s - loss: 0.2489 - mean_squared_error: 0.0624 - acc: 0.9377 - val_loss: 0.3392 - val_mean_squared_error: 0.0998 - val_acc: 0.8650\n",
      "Epoch 505/3000\n",
      " - 0s - loss: 0.2482 - mean_squared_error: 0.0623 - acc: 0.9425 - val_loss: 0.3390 - val_mean_squared_error: 0.0998 - val_acc: 0.8650\n",
      "Epoch 506/3000\n",
      " - 0s - loss: 0.2471 - mean_squared_error: 0.0622 - acc: 0.9425 - val_loss: 0.3389 - val_mean_squared_error: 0.0997 - val_acc: 0.8650\n",
      "Epoch 507/3000\n",
      " - 0s - loss: 0.2495 - mean_squared_error: 0.0631 - acc: 0.9401 - val_loss: 0.3387 - val_mean_squared_error: 0.0997 - val_acc: 0.8650\n",
      "Epoch 508/3000\n",
      " - 0s - loss: 0.2454 - mean_squared_error: 0.0616 - acc: 0.9401 - val_loss: 0.3385 - val_mean_squared_error: 0.0996 - val_acc: 0.8650\n",
      "Epoch 509/3000\n",
      " - 0s - loss: 0.2443 - mean_squared_error: 0.0609 - acc: 0.9389 - val_loss: 0.3382 - val_mean_squared_error: 0.0995 - val_acc: 0.8650\n",
      "Epoch 510/3000\n",
      " - 0s - loss: 0.2436 - mean_squared_error: 0.0606 - acc: 0.9425 - val_loss: 0.3380 - val_mean_squared_error: 0.0995 - val_acc: 0.8650\n",
      "Epoch 511/3000\n",
      " - 0s - loss: 0.2484 - mean_squared_error: 0.0625 - acc: 0.9438 - val_loss: 0.3377 - val_mean_squared_error: 0.0994 - val_acc: 0.8650\n",
      "Epoch 512/3000\n",
      " - 0s - loss: 0.2459 - mean_squared_error: 0.0619 - acc: 0.9340 - val_loss: 0.3374 - val_mean_squared_error: 0.0993 - val_acc: 0.8650\n",
      "Epoch 513/3000\n",
      " - 0s - loss: 0.2447 - mean_squared_error: 0.0614 - acc: 0.9425 - val_loss: 0.3372 - val_mean_squared_error: 0.0993 - val_acc: 0.8650\n",
      "Epoch 514/3000\n",
      " - 0s - loss: 0.2496 - mean_squared_error: 0.0634 - acc: 0.9352 - val_loss: 0.3370 - val_mean_squared_error: 0.0992 - val_acc: 0.8650\n",
      "Epoch 515/3000\n",
      " - 0s - loss: 0.2463 - mean_squared_error: 0.0621 - acc: 0.9413 - val_loss: 0.3368 - val_mean_squared_error: 0.0992 - val_acc: 0.8650\n",
      "Epoch 516/3000\n",
      " - 0s - loss: 0.2440 - mean_squared_error: 0.0613 - acc: 0.9450 - val_loss: 0.3367 - val_mean_squared_error: 0.0991 - val_acc: 0.8650\n",
      "Epoch 517/3000\n",
      " - 0s - loss: 0.2413 - mean_squared_error: 0.0605 - acc: 0.9377 - val_loss: 0.3366 - val_mean_squared_error: 0.0991 - val_acc: 0.8650\n",
      "Epoch 518/3000\n",
      " - 0s - loss: 0.2467 - mean_squared_error: 0.0624 - acc: 0.9401 - val_loss: 0.3365 - val_mean_squared_error: 0.0991 - val_acc: 0.8650\n",
      "Epoch 519/3000\n",
      " - 0s - loss: 0.2411 - mean_squared_error: 0.0601 - acc: 0.9438 - val_loss: 0.3364 - val_mean_squared_error: 0.0991 - val_acc: 0.8650\n",
      "Epoch 520/3000\n",
      " - 0s - loss: 0.2423 - mean_squared_error: 0.0606 - acc: 0.9352 - val_loss: 0.3364 - val_mean_squared_error: 0.0991 - val_acc: 0.8650\n",
      "Epoch 521/3000\n",
      " - 0s - loss: 0.2428 - mean_squared_error: 0.0610 - acc: 0.9425 - val_loss: 0.3363 - val_mean_squared_error: 0.0990 - val_acc: 0.8650\n",
      "Epoch 522/3000\n",
      " - 0s - loss: 0.2419 - mean_squared_error: 0.0610 - acc: 0.9364 - val_loss: 0.3363 - val_mean_squared_error: 0.0990 - val_acc: 0.8650\n",
      "Epoch 523/3000\n",
      " - 0s - loss: 0.2458 - mean_squared_error: 0.0620 - acc: 0.9377 - val_loss: 0.3362 - val_mean_squared_error: 0.0990 - val_acc: 0.8650\n",
      "Epoch 524/3000\n",
      " - 0s - loss: 0.2402 - mean_squared_error: 0.0599 - acc: 0.9389 - val_loss: 0.3361 - val_mean_squared_error: 0.0990 - val_acc: 0.8650\n",
      "Epoch 525/3000\n",
      " - 0s - loss: 0.2412 - mean_squared_error: 0.0605 - acc: 0.9401 - val_loss: 0.3359 - val_mean_squared_error: 0.0989 - val_acc: 0.8650\n",
      "Epoch 526/3000\n",
      " - 0s - loss: 0.2435 - mean_squared_error: 0.0614 - acc: 0.9364 - val_loss: 0.3358 - val_mean_squared_error: 0.0989 - val_acc: 0.8650\n",
      "Epoch 527/3000\n",
      " - 0s - loss: 0.2448 - mean_squared_error: 0.0614 - acc: 0.9364 - val_loss: 0.3356 - val_mean_squared_error: 0.0989 - val_acc: 0.8650\n",
      "Epoch 528/3000\n",
      " - 0s - loss: 0.2403 - mean_squared_error: 0.0604 - acc: 0.9377 - val_loss: 0.3354 - val_mean_squared_error: 0.0988 - val_acc: 0.8650\n",
      "Epoch 529/3000\n",
      " - 0s - loss: 0.2426 - mean_squared_error: 0.0608 - acc: 0.9413 - val_loss: 0.3352 - val_mean_squared_error: 0.0987 - val_acc: 0.8650\n",
      "Epoch 530/3000\n",
      " - 0s - loss: 0.2356 - mean_squared_error: 0.0584 - acc: 0.9474 - val_loss: 0.3349 - val_mean_squared_error: 0.0987 - val_acc: 0.8650\n",
      "Epoch 531/3000\n",
      " - 0s - loss: 0.2412 - mean_squared_error: 0.0605 - acc: 0.9377 - val_loss: 0.3346 - val_mean_squared_error: 0.0986 - val_acc: 0.8650\n",
      "Epoch 532/3000\n",
      " - 0s - loss: 0.2431 - mean_squared_error: 0.0616 - acc: 0.9315 - val_loss: 0.3344 - val_mean_squared_error: 0.0985 - val_acc: 0.8650\n",
      "Epoch 533/3000\n",
      " - 0s - loss: 0.2429 - mean_squared_error: 0.0616 - acc: 0.9340 - val_loss: 0.3342 - val_mean_squared_error: 0.0985 - val_acc: 0.8650\n",
      "Epoch 534/3000\n",
      " - 0s - loss: 0.2388 - mean_squared_error: 0.0594 - acc: 0.9413 - val_loss: 0.3340 - val_mean_squared_error: 0.0984 - val_acc: 0.8650\n",
      "Epoch 535/3000\n",
      " - 0s - loss: 0.2386 - mean_squared_error: 0.0600 - acc: 0.9413 - val_loss: 0.3338 - val_mean_squared_error: 0.0984 - val_acc: 0.8650\n",
      "Epoch 536/3000\n",
      " - 0s - loss: 0.2405 - mean_squared_error: 0.0602 - acc: 0.9438 - val_loss: 0.3337 - val_mean_squared_error: 0.0984 - val_acc: 0.8650\n",
      "Epoch 537/3000\n",
      " - 0s - loss: 0.2375 - mean_squared_error: 0.0590 - acc: 0.9389 - val_loss: 0.3336 - val_mean_squared_error: 0.0983 - val_acc: 0.8650\n",
      "Epoch 538/3000\n",
      " - 0s - loss: 0.2398 - mean_squared_error: 0.0605 - acc: 0.9401 - val_loss: 0.3335 - val_mean_squared_error: 0.0983 - val_acc: 0.8650\n",
      "Epoch 539/3000\n",
      " - 0s - loss: 0.2395 - mean_squared_error: 0.0601 - acc: 0.9389 - val_loss: 0.3334 - val_mean_squared_error: 0.0983 - val_acc: 0.8650\n",
      "Epoch 540/3000\n",
      " - 0s - loss: 0.2371 - mean_squared_error: 0.0592 - acc: 0.9328 - val_loss: 0.3334 - val_mean_squared_error: 0.0983 - val_acc: 0.8650\n",
      "Epoch 541/3000\n",
      " - 0s - loss: 0.2343 - mean_squared_error: 0.0580 - acc: 0.9450 - val_loss: 0.3334 - val_mean_squared_error: 0.0983 - val_acc: 0.8650\n",
      "Epoch 542/3000\n",
      " - 0s - loss: 0.2363 - mean_squared_error: 0.0589 - acc: 0.9425 - val_loss: 0.3333 - val_mean_squared_error: 0.0983 - val_acc: 0.8650\n",
      "Epoch 543/3000\n",
      " - 0s - loss: 0.2327 - mean_squared_error: 0.0575 - acc: 0.9413 - val_loss: 0.3332 - val_mean_squared_error: 0.0982 - val_acc: 0.8650\n",
      "Epoch 544/3000\n",
      " - 0s - loss: 0.2347 - mean_squared_error: 0.0581 - acc: 0.9401 - val_loss: 0.3331 - val_mean_squared_error: 0.0982 - val_acc: 0.8650\n",
      "Epoch 545/3000\n",
      " - 0s - loss: 0.2396 - mean_squared_error: 0.0596 - acc: 0.9462 - val_loss: 0.3330 - val_mean_squared_error: 0.0982 - val_acc: 0.8650\n",
      "Epoch 546/3000\n",
      " - 0s - loss: 0.2332 - mean_squared_error: 0.0577 - acc: 0.9377 - val_loss: 0.3329 - val_mean_squared_error: 0.0982 - val_acc: 0.8650\n",
      "Epoch 547/3000\n",
      " - 0s - loss: 0.2330 - mean_squared_error: 0.0572 - acc: 0.9487 - val_loss: 0.3327 - val_mean_squared_error: 0.0981 - val_acc: 0.8650\n",
      "Epoch 548/3000\n",
      " - 0s - loss: 0.2373 - mean_squared_error: 0.0594 - acc: 0.9401 - val_loss: 0.3325 - val_mean_squared_error: 0.0981 - val_acc: 0.8650\n",
      "Epoch 549/3000\n",
      " - 0s - loss: 0.2340 - mean_squared_error: 0.0581 - acc: 0.9413 - val_loss: 0.3323 - val_mean_squared_error: 0.0980 - val_acc: 0.8650\n",
      "Epoch 550/3000\n",
      " - 0s - loss: 0.2346 - mean_squared_error: 0.0588 - acc: 0.9401 - val_loss: 0.3321 - val_mean_squared_error: 0.0980 - val_acc: 0.8650\n",
      "Epoch 551/3000\n",
      " - 0s - loss: 0.2362 - mean_squared_error: 0.0589 - acc: 0.9377 - val_loss: 0.3318 - val_mean_squared_error: 0.0979 - val_acc: 0.8650\n",
      "Epoch 552/3000\n",
      " - 0s - loss: 0.2295 - mean_squared_error: 0.0569 - acc: 0.9389 - val_loss: 0.3316 - val_mean_squared_error: 0.0978 - val_acc: 0.8650\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 553/3000\n",
      " - 0s - loss: 0.2374 - mean_squared_error: 0.0594 - acc: 0.9352 - val_loss: 0.3314 - val_mean_squared_error: 0.0978 - val_acc: 0.8650\n",
      "Epoch 554/3000\n",
      " - 0s - loss: 0.2345 - mean_squared_error: 0.0587 - acc: 0.9352 - val_loss: 0.3313 - val_mean_squared_error: 0.0978 - val_acc: 0.8650\n",
      "Epoch 555/3000\n",
      " - 0s - loss: 0.2311 - mean_squared_error: 0.0572 - acc: 0.9413 - val_loss: 0.3311 - val_mean_squared_error: 0.0977 - val_acc: 0.8650\n",
      "Epoch 556/3000\n",
      " - 0s - loss: 0.2334 - mean_squared_error: 0.0580 - acc: 0.9401 - val_loss: 0.3310 - val_mean_squared_error: 0.0977 - val_acc: 0.8650\n",
      "Epoch 557/3000\n",
      " - 0s - loss: 0.2329 - mean_squared_error: 0.0576 - acc: 0.9450 - val_loss: 0.3309 - val_mean_squared_error: 0.0977 - val_acc: 0.8650\n",
      "Epoch 558/3000\n",
      " - 0s - loss: 0.2337 - mean_squared_error: 0.0576 - acc: 0.9438 - val_loss: 0.3308 - val_mean_squared_error: 0.0976 - val_acc: 0.8650\n",
      "Epoch 559/3000\n",
      " - 0s - loss: 0.2289 - mean_squared_error: 0.0561 - acc: 0.9438 - val_loss: 0.3307 - val_mean_squared_error: 0.0976 - val_acc: 0.8650\n",
      "Epoch 560/3000\n",
      " - 0s - loss: 0.2294 - mean_squared_error: 0.0568 - acc: 0.9401 - val_loss: 0.3307 - val_mean_squared_error: 0.0976 - val_acc: 0.8650\n",
      "Epoch 561/3000\n",
      " - 0s - loss: 0.2296 - mean_squared_error: 0.0566 - acc: 0.9462 - val_loss: 0.3307 - val_mean_squared_error: 0.0976 - val_acc: 0.8650\n",
      "Epoch 562/3000\n",
      " - 0s - loss: 0.2266 - mean_squared_error: 0.0557 - acc: 0.9487 - val_loss: 0.3306 - val_mean_squared_error: 0.0976 - val_acc: 0.8650\n",
      "Epoch 563/3000\n",
      " - 0s - loss: 0.2269 - mean_squared_error: 0.0552 - acc: 0.9511 - val_loss: 0.3305 - val_mean_squared_error: 0.0976 - val_acc: 0.8650\n",
      "Epoch 564/3000\n",
      " - 0s - loss: 0.2300 - mean_squared_error: 0.0575 - acc: 0.9364 - val_loss: 0.3305 - val_mean_squared_error: 0.0975 - val_acc: 0.8650\n",
      "Epoch 565/3000\n",
      " - 0s - loss: 0.2357 - mean_squared_error: 0.0589 - acc: 0.9401 - val_loss: 0.3303 - val_mean_squared_error: 0.0975 - val_acc: 0.8650\n",
      "Epoch 566/3000\n",
      " - 0s - loss: 0.2357 - mean_squared_error: 0.0594 - acc: 0.9401 - val_loss: 0.3302 - val_mean_squared_error: 0.0975 - val_acc: 0.8650\n",
      "Epoch 567/3000\n",
      " - 0s - loss: 0.2270 - mean_squared_error: 0.0558 - acc: 0.9413 - val_loss: 0.3301 - val_mean_squared_error: 0.0975 - val_acc: 0.8650\n",
      "Epoch 568/3000\n",
      " - 0s - loss: 0.2274 - mean_squared_error: 0.0555 - acc: 0.9462 - val_loss: 0.3299 - val_mean_squared_error: 0.0974 - val_acc: 0.8650\n",
      "Epoch 569/3000\n",
      " - 0s - loss: 0.2268 - mean_squared_error: 0.0557 - acc: 0.9425 - val_loss: 0.3297 - val_mean_squared_error: 0.0974 - val_acc: 0.8650\n",
      "Epoch 570/3000\n",
      " - 0s - loss: 0.2245 - mean_squared_error: 0.0551 - acc: 0.9462 - val_loss: 0.3295 - val_mean_squared_error: 0.0973 - val_acc: 0.8650\n",
      "Epoch 571/3000\n",
      " - 0s - loss: 0.2287 - mean_squared_error: 0.0566 - acc: 0.9450 - val_loss: 0.3293 - val_mean_squared_error: 0.0973 - val_acc: 0.8650\n",
      "Epoch 572/3000\n",
      " - 0s - loss: 0.2270 - mean_squared_error: 0.0561 - acc: 0.9438 - val_loss: 0.3291 - val_mean_squared_error: 0.0972 - val_acc: 0.8650\n",
      "Epoch 573/3000\n",
      " - 0s - loss: 0.2258 - mean_squared_error: 0.0555 - acc: 0.9413 - val_loss: 0.3289 - val_mean_squared_error: 0.0972 - val_acc: 0.8650\n",
      "Epoch 574/3000\n",
      " - 0s - loss: 0.2293 - mean_squared_error: 0.0567 - acc: 0.9401 - val_loss: 0.3287 - val_mean_squared_error: 0.0971 - val_acc: 0.8650\n",
      "Epoch 575/3000\n",
      " - 0s - loss: 0.2272 - mean_squared_error: 0.0558 - acc: 0.9401 - val_loss: 0.3286 - val_mean_squared_error: 0.0971 - val_acc: 0.8650\n",
      "Epoch 576/3000\n",
      " - 0s - loss: 0.2236 - mean_squared_error: 0.0547 - acc: 0.9511 - val_loss: 0.3285 - val_mean_squared_error: 0.0971 - val_acc: 0.8650\n",
      "Epoch 577/3000\n",
      " - 0s - loss: 0.2295 - mean_squared_error: 0.0570 - acc: 0.9438 - val_loss: 0.3284 - val_mean_squared_error: 0.0970 - val_acc: 0.8650\n",
      "Epoch 578/3000\n",
      " - 0s - loss: 0.2230 - mean_squared_error: 0.0548 - acc: 0.9389 - val_loss: 0.3283 - val_mean_squared_error: 0.0970 - val_acc: 0.8650\n",
      "Epoch 579/3000\n",
      " - 0s - loss: 0.2280 - mean_squared_error: 0.0566 - acc: 0.9413 - val_loss: 0.3282 - val_mean_squared_error: 0.0970 - val_acc: 0.8650\n",
      "Epoch 580/3000\n",
      " - 0s - loss: 0.2230 - mean_squared_error: 0.0547 - acc: 0.9474 - val_loss: 0.3282 - val_mean_squared_error: 0.0970 - val_acc: 0.8650\n",
      "Epoch 581/3000\n",
      " - 0s - loss: 0.2223 - mean_squared_error: 0.0544 - acc: 0.9425 - val_loss: 0.3282 - val_mean_squared_error: 0.0970 - val_acc: 0.8650\n",
      "Epoch 582/3000\n",
      " - 0s - loss: 0.2246 - mean_squared_error: 0.0555 - acc: 0.9413 - val_loss: 0.3281 - val_mean_squared_error: 0.0970 - val_acc: 0.8650\n",
      "Epoch 583/3000\n",
      " - 0s - loss: 0.2219 - mean_squared_error: 0.0545 - acc: 0.9377 - val_loss: 0.3281 - val_mean_squared_error: 0.0970 - val_acc: 0.8650\n",
      "Epoch 584/3000\n",
      " - 0s - loss: 0.2225 - mean_squared_error: 0.0547 - acc: 0.9450 - val_loss: 0.3280 - val_mean_squared_error: 0.0970 - val_acc: 0.8650\n",
      "Epoch 585/3000\n",
      " - 0s - loss: 0.2227 - mean_squared_error: 0.0547 - acc: 0.9401 - val_loss: 0.3279 - val_mean_squared_error: 0.0969 - val_acc: 0.8650\n",
      "Epoch 586/3000\n",
      " - 0s - loss: 0.2260 - mean_squared_error: 0.0555 - acc: 0.9425 - val_loss: 0.3277 - val_mean_squared_error: 0.0969 - val_acc: 0.8650\n",
      "Epoch 587/3000\n",
      " - 0s - loss: 0.2194 - mean_squared_error: 0.0533 - acc: 0.9487 - val_loss: 0.3276 - val_mean_squared_error: 0.0969 - val_acc: 0.8650\n",
      "Epoch 588/3000\n",
      " - 0s - loss: 0.2259 - mean_squared_error: 0.0557 - acc: 0.9425 - val_loss: 0.3274 - val_mean_squared_error: 0.0968 - val_acc: 0.8650\n",
      "Epoch 589/3000\n",
      " - 0s - loss: 0.2222 - mean_squared_error: 0.0543 - acc: 0.9499 - val_loss: 0.3273 - val_mean_squared_error: 0.0968 - val_acc: 0.8650\n",
      "Epoch 590/3000\n",
      " - 0s - loss: 0.2237 - mean_squared_error: 0.0550 - acc: 0.9450 - val_loss: 0.3271 - val_mean_squared_error: 0.0967 - val_acc: 0.8650\n",
      "Epoch 591/3000\n",
      " - 0s - loss: 0.2236 - mean_squared_error: 0.0553 - acc: 0.9462 - val_loss: 0.3269 - val_mean_squared_error: 0.0967 - val_acc: 0.8650\n",
      "Epoch 592/3000\n",
      " - 0s - loss: 0.2217 - mean_squared_error: 0.0538 - acc: 0.9487 - val_loss: 0.3267 - val_mean_squared_error: 0.0966 - val_acc: 0.8650\n",
      "Epoch 593/3000\n",
      " - 0s - loss: 0.2237 - mean_squared_error: 0.0552 - acc: 0.9425 - val_loss: 0.3265 - val_mean_squared_error: 0.0966 - val_acc: 0.8650\n",
      "Epoch 594/3000\n",
      " - 0s - loss: 0.2173 - mean_squared_error: 0.0531 - acc: 0.9377 - val_loss: 0.3264 - val_mean_squared_error: 0.0966 - val_acc: 0.8650\n",
      "Epoch 595/3000\n",
      " - 0s - loss: 0.2224 - mean_squared_error: 0.0549 - acc: 0.9425 - val_loss: 0.3262 - val_mean_squared_error: 0.0965 - val_acc: 0.8650\n",
      "Epoch 596/3000\n",
      " - 0s - loss: 0.2175 - mean_squared_error: 0.0527 - acc: 0.9535 - val_loss: 0.3261 - val_mean_squared_error: 0.0965 - val_acc: 0.8650\n",
      "Epoch 597/3000\n",
      " - 0s - loss: 0.2198 - mean_squared_error: 0.0541 - acc: 0.9499 - val_loss: 0.3260 - val_mean_squared_error: 0.0965 - val_acc: 0.8650\n",
      "Epoch 598/3000\n",
      " - 0s - loss: 0.2196 - mean_squared_error: 0.0538 - acc: 0.9462 - val_loss: 0.3260 - val_mean_squared_error: 0.0965 - val_acc: 0.8650\n",
      "Epoch 599/3000\n",
      " - 0s - loss: 0.2214 - mean_squared_error: 0.0546 - acc: 0.9389 - val_loss: 0.3259 - val_mean_squared_error: 0.0965 - val_acc: 0.8650\n",
      "Epoch 600/3000\n",
      " - 0s - loss: 0.2219 - mean_squared_error: 0.0543 - acc: 0.9438 - val_loss: 0.3259 - val_mean_squared_error: 0.0965 - val_acc: 0.8650\n",
      "Epoch 601/3000\n",
      " - 0s - loss: 0.2172 - mean_squared_error: 0.0528 - acc: 0.9474 - val_loss: 0.3259 - val_mean_squared_error: 0.0965 - val_acc: 0.8650\n",
      "Epoch 602/3000\n",
      " - 0s - loss: 0.2183 - mean_squared_error: 0.0534 - acc: 0.9462 - val_loss: 0.3258 - val_mean_squared_error: 0.0964 - val_acc: 0.8650\n",
      "Epoch 603/3000\n",
      " - 0s - loss: 0.2203 - mean_squared_error: 0.0538 - acc: 0.9474 - val_loss: 0.3258 - val_mean_squared_error: 0.0964 - val_acc: 0.8650\n",
      "Epoch 604/3000\n",
      " - 0s - loss: 0.2182 - mean_squared_error: 0.0536 - acc: 0.9425 - val_loss: 0.3257 - val_mean_squared_error: 0.0964 - val_acc: 0.8650\n",
      "Epoch 605/3000\n",
      " - 0s - loss: 0.2153 - mean_squared_error: 0.0517 - acc: 0.9474 - val_loss: 0.3256 - val_mean_squared_error: 0.0964 - val_acc: 0.8650\n",
      "Epoch 606/3000\n",
      " - 0s - loss: 0.2168 - mean_squared_error: 0.0528 - acc: 0.9462 - val_loss: 0.3255 - val_mean_squared_error: 0.0964 - val_acc: 0.8650\n",
      "Epoch 607/3000\n",
      " - 0s - loss: 0.2179 - mean_squared_error: 0.0533 - acc: 0.9487 - val_loss: 0.3253 - val_mean_squared_error: 0.0963 - val_acc: 0.8650\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 608/3000\n",
      " - 0s - loss: 0.2185 - mean_squared_error: 0.0531 - acc: 0.9450 - val_loss: 0.3252 - val_mean_squared_error: 0.0963 - val_acc: 0.8650\n",
      "Epoch 609/3000\n",
      " - 0s - loss: 0.2203 - mean_squared_error: 0.0543 - acc: 0.9364 - val_loss: 0.3250 - val_mean_squared_error: 0.0963 - val_acc: 0.8650\n",
      "Epoch 610/3000\n",
      " - 0s - loss: 0.2193 - mean_squared_error: 0.0537 - acc: 0.9438 - val_loss: 0.3249 - val_mean_squared_error: 0.0962 - val_acc: 0.8650\n",
      "Epoch 611/3000\n",
      " - 0s - loss: 0.2153 - mean_squared_error: 0.0523 - acc: 0.9450 - val_loss: 0.3247 - val_mean_squared_error: 0.0962 - val_acc: 0.8650\n",
      "Epoch 612/3000\n",
      " - 0s - loss: 0.2141 - mean_squared_error: 0.0520 - acc: 0.9438 - val_loss: 0.3245 - val_mean_squared_error: 0.0961 - val_acc: 0.8650\n",
      "Epoch 613/3000\n",
      " - 0s - loss: 0.2205 - mean_squared_error: 0.0544 - acc: 0.9499 - val_loss: 0.3243 - val_mean_squared_error: 0.0961 - val_acc: 0.8650\n",
      "Epoch 614/3000\n",
      " - 0s - loss: 0.2134 - mean_squared_error: 0.0513 - acc: 0.9474 - val_loss: 0.3242 - val_mean_squared_error: 0.0961 - val_acc: 0.8650\n",
      "Epoch 615/3000\n",
      " - 0s - loss: 0.2154 - mean_squared_error: 0.0524 - acc: 0.9462 - val_loss: 0.3241 - val_mean_squared_error: 0.0961 - val_acc: 0.8650\n",
      "Epoch 616/3000\n",
      " - 0s - loss: 0.2143 - mean_squared_error: 0.0521 - acc: 0.9487 - val_loss: 0.3240 - val_mean_squared_error: 0.0960 - val_acc: 0.8650\n",
      "Epoch 617/3000\n",
      " - 0s - loss: 0.2129 - mean_squared_error: 0.0515 - acc: 0.9438 - val_loss: 0.3239 - val_mean_squared_error: 0.0960 - val_acc: 0.8650\n",
      "Epoch 618/3000\n",
      " - 0s - loss: 0.2148 - mean_squared_error: 0.0520 - acc: 0.9487 - val_loss: 0.3239 - val_mean_squared_error: 0.0960 - val_acc: 0.8650\n",
      "Epoch 619/3000\n",
      " - 0s - loss: 0.2142 - mean_squared_error: 0.0522 - acc: 0.9364 - val_loss: 0.3238 - val_mean_squared_error: 0.0960 - val_acc: 0.8650\n",
      "Epoch 620/3000\n",
      " - 0s - loss: 0.2142 - mean_squared_error: 0.0521 - acc: 0.9413 - val_loss: 0.3238 - val_mean_squared_error: 0.0960 - val_acc: 0.8650\n",
      "Epoch 621/3000\n",
      " - 0s - loss: 0.2170 - mean_squared_error: 0.0531 - acc: 0.9438 - val_loss: 0.3238 - val_mean_squared_error: 0.0960 - val_acc: 0.8650\n",
      "Epoch 622/3000\n",
      " - 0s - loss: 0.2134 - mean_squared_error: 0.0516 - acc: 0.9438 - val_loss: 0.3237 - val_mean_squared_error: 0.0960 - val_acc: 0.8650\n",
      "Epoch 623/3000\n",
      " - 0s - loss: 0.2112 - mean_squared_error: 0.0508 - acc: 0.9499 - val_loss: 0.3237 - val_mean_squared_error: 0.0960 - val_acc: 0.8650\n",
      "Epoch 624/3000\n",
      " - 0s - loss: 0.2178 - mean_squared_error: 0.0536 - acc: 0.9425 - val_loss: 0.3236 - val_mean_squared_error: 0.0959 - val_acc: 0.8650\n",
      "Epoch 625/3000\n",
      " - 0s - loss: 0.2102 - mean_squared_error: 0.0506 - acc: 0.9499 - val_loss: 0.3235 - val_mean_squared_error: 0.0959 - val_acc: 0.8650\n",
      "Epoch 626/3000\n",
      " - 0s - loss: 0.2181 - mean_squared_error: 0.0538 - acc: 0.9474 - val_loss: 0.3234 - val_mean_squared_error: 0.0959 - val_acc: 0.8650\n",
      "Epoch 627/3000\n",
      " - 0s - loss: 0.2166 - mean_squared_error: 0.0525 - acc: 0.9499 - val_loss: 0.3233 - val_mean_squared_error: 0.0959 - val_acc: 0.8650\n",
      "Epoch 628/3000\n",
      " - 0s - loss: 0.2149 - mean_squared_error: 0.0526 - acc: 0.9450 - val_loss: 0.3231 - val_mean_squared_error: 0.0959 - val_acc: 0.8650\n",
      "Epoch 629/3000\n",
      " - 0s - loss: 0.2154 - mean_squared_error: 0.0527 - acc: 0.9450 - val_loss: 0.3230 - val_mean_squared_error: 0.0958 - val_acc: 0.8650\n",
      "Epoch 630/3000\n",
      " - 0s - loss: 0.2179 - mean_squared_error: 0.0535 - acc: 0.9438 - val_loss: 0.3228 - val_mean_squared_error: 0.0958 - val_acc: 0.8650\n",
      "Epoch 631/3000\n",
      " - 0s - loss: 0.2153 - mean_squared_error: 0.0532 - acc: 0.9401 - val_loss: 0.3227 - val_mean_squared_error: 0.0957 - val_acc: 0.8650\n",
      "Epoch 632/3000\n",
      " - 0s - loss: 0.2112 - mean_squared_error: 0.0517 - acc: 0.9401 - val_loss: 0.3225 - val_mean_squared_error: 0.0957 - val_acc: 0.8650\n",
      "Epoch 633/3000\n",
      " - 0s - loss: 0.2099 - mean_squared_error: 0.0509 - acc: 0.9487 - val_loss: 0.3224 - val_mean_squared_error: 0.0957 - val_acc: 0.8650\n",
      "Epoch 634/3000\n",
      " - 0s - loss: 0.2125 - mean_squared_error: 0.0518 - acc: 0.9487 - val_loss: 0.3222 - val_mean_squared_error: 0.0957 - val_acc: 0.8650\n",
      "Epoch 635/3000\n",
      " - 0s - loss: 0.2113 - mean_squared_error: 0.0507 - acc: 0.9511 - val_loss: 0.3221 - val_mean_squared_error: 0.0956 - val_acc: 0.8650\n",
      "Epoch 636/3000\n",
      " - 0s - loss: 0.2129 - mean_squared_error: 0.0521 - acc: 0.9462 - val_loss: 0.3221 - val_mean_squared_error: 0.0956 - val_acc: 0.8650\n",
      "Epoch 637/3000\n",
      " - 0s - loss: 0.2062 - mean_squared_error: 0.0498 - acc: 0.9462 - val_loss: 0.3220 - val_mean_squared_error: 0.0956 - val_acc: 0.8650\n",
      "Epoch 638/3000\n",
      " - 0s - loss: 0.2138 - mean_squared_error: 0.0523 - acc: 0.9425 - val_loss: 0.3219 - val_mean_squared_error: 0.0956 - val_acc: 0.8650\n",
      "Epoch 639/3000\n",
      " - 0s - loss: 0.2137 - mean_squared_error: 0.0525 - acc: 0.9438 - val_loss: 0.3219 - val_mean_squared_error: 0.0956 - val_acc: 0.8650\n",
      "Epoch 640/3000\n",
      " - 0s - loss: 0.2103 - mean_squared_error: 0.0508 - acc: 0.9487 - val_loss: 0.3218 - val_mean_squared_error: 0.0956 - val_acc: 0.8650\n",
      "Epoch 641/3000\n",
      " - 0s - loss: 0.2044 - mean_squared_error: 0.0490 - acc: 0.9462 - val_loss: 0.3218 - val_mean_squared_error: 0.0956 - val_acc: 0.8650\n",
      "Epoch 642/3000\n",
      " - 0s - loss: 0.2095 - mean_squared_error: 0.0509 - acc: 0.9450 - val_loss: 0.3218 - val_mean_squared_error: 0.0956 - val_acc: 0.8650\n",
      "Epoch 643/3000\n",
      " - 0s - loss: 0.2123 - mean_squared_error: 0.0521 - acc: 0.9462 - val_loss: 0.3217 - val_mean_squared_error: 0.0955 - val_acc: 0.8650\n",
      "Epoch 644/3000\n",
      " - 0s - loss: 0.2083 - mean_squared_error: 0.0504 - acc: 0.9487 - val_loss: 0.3217 - val_mean_squared_error: 0.0955 - val_acc: 0.8650\n",
      "Epoch 645/3000\n",
      " - 0s - loss: 0.2094 - mean_squared_error: 0.0505 - acc: 0.9462 - val_loss: 0.3216 - val_mean_squared_error: 0.0955 - val_acc: 0.8650\n",
      "Epoch 646/3000\n",
      " - 0s - loss: 0.2106 - mean_squared_error: 0.0509 - acc: 0.9499 - val_loss: 0.3215 - val_mean_squared_error: 0.0955 - val_acc: 0.8650\n",
      "Epoch 647/3000\n",
      " - 0s - loss: 0.2057 - mean_squared_error: 0.0493 - acc: 0.9499 - val_loss: 0.3214 - val_mean_squared_error: 0.0955 - val_acc: 0.8650\n",
      "Epoch 648/3000\n",
      " - 0s - loss: 0.2111 - mean_squared_error: 0.0515 - acc: 0.9425 - val_loss: 0.3213 - val_mean_squared_error: 0.0954 - val_acc: 0.8650\n",
      "Epoch 649/3000\n",
      " - 0s - loss: 0.2090 - mean_squared_error: 0.0500 - acc: 0.9523 - val_loss: 0.3211 - val_mean_squared_error: 0.0954 - val_acc: 0.8650\n",
      "Epoch 650/3000\n",
      " - 0s - loss: 0.2053 - mean_squared_error: 0.0492 - acc: 0.9560 - val_loss: 0.3210 - val_mean_squared_error: 0.0954 - val_acc: 0.8650\n",
      "Epoch 651/3000\n",
      " - 0s - loss: 0.2036 - mean_squared_error: 0.0485 - acc: 0.9474 - val_loss: 0.3208 - val_mean_squared_error: 0.0954 - val_acc: 0.8650\n",
      "Epoch 652/3000\n",
      " - 0s - loss: 0.2105 - mean_squared_error: 0.0514 - acc: 0.9487 - val_loss: 0.3207 - val_mean_squared_error: 0.0953 - val_acc: 0.8650\n",
      "Epoch 653/3000\n",
      " - 0s - loss: 0.2076 - mean_squared_error: 0.0506 - acc: 0.9450 - val_loss: 0.3206 - val_mean_squared_error: 0.0953 - val_acc: 0.8650\n",
      "Epoch 654/3000\n",
      " - 0s - loss: 0.2071 - mean_squared_error: 0.0500 - acc: 0.9487 - val_loss: 0.3204 - val_mean_squared_error: 0.0953 - val_acc: 0.8650\n",
      "Epoch 655/3000\n",
      " - 0s - loss: 0.2031 - mean_squared_error: 0.0488 - acc: 0.9487 - val_loss: 0.3203 - val_mean_squared_error: 0.0953 - val_acc: 0.8650\n",
      "Epoch 656/3000\n",
      " - 0s - loss: 0.2083 - mean_squared_error: 0.0510 - acc: 0.9413 - val_loss: 0.3203 - val_mean_squared_error: 0.0952 - val_acc: 0.8650\n",
      "Epoch 657/3000\n",
      " - 0s - loss: 0.2059 - mean_squared_error: 0.0496 - acc: 0.9450 - val_loss: 0.3202 - val_mean_squared_error: 0.0952 - val_acc: 0.8650\n",
      "Epoch 658/3000\n",
      " - 0s - loss: 0.2101 - mean_squared_error: 0.0517 - acc: 0.9413 - val_loss: 0.3201 - val_mean_squared_error: 0.0952 - val_acc: 0.8650\n",
      "Epoch 659/3000\n",
      " - 0s - loss: 0.2074 - mean_squared_error: 0.0505 - acc: 0.9462 - val_loss: 0.3201 - val_mean_squared_error: 0.0952 - val_acc: 0.8650\n",
      "Epoch 660/3000\n",
      " - 0s - loss: 0.2052 - mean_squared_error: 0.0494 - acc: 0.9487 - val_loss: 0.3201 - val_mean_squared_error: 0.0952 - val_acc: 0.8650\n",
      "Epoch 661/3000\n",
      " - 0s - loss: 0.2024 - mean_squared_error: 0.0478 - acc: 0.9523 - val_loss: 0.3201 - val_mean_squared_error: 0.0952 - val_acc: 0.8650\n",
      "Epoch 662/3000\n",
      " - 0s - loss: 0.2069 - mean_squared_error: 0.0498 - acc: 0.9499 - val_loss: 0.3200 - val_mean_squared_error: 0.0952 - val_acc: 0.8650\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 663/3000\n",
      " - 0s - loss: 0.2024 - mean_squared_error: 0.0482 - acc: 0.9523 - val_loss: 0.3200 - val_mean_squared_error: 0.0952 - val_acc: 0.8650\n",
      "Epoch 664/3000\n",
      " - 0s - loss: 0.2059 - mean_squared_error: 0.0494 - acc: 0.9462 - val_loss: 0.3199 - val_mean_squared_error: 0.0952 - val_acc: 0.8650\n",
      "Epoch 665/3000\n",
      " - 0s - loss: 0.2022 - mean_squared_error: 0.0484 - acc: 0.9511 - val_loss: 0.3198 - val_mean_squared_error: 0.0952 - val_acc: 0.8650\n",
      "Epoch 666/3000\n",
      " - 0s - loss: 0.2004 - mean_squared_error: 0.0477 - acc: 0.9511 - val_loss: 0.3197 - val_mean_squared_error: 0.0951 - val_acc: 0.8650\n",
      "Epoch 667/3000\n",
      " - 0s - loss: 0.2020 - mean_squared_error: 0.0483 - acc: 0.9499 - val_loss: 0.3196 - val_mean_squared_error: 0.0951 - val_acc: 0.8650\n",
      "Epoch 668/3000\n",
      " - 0s - loss: 0.2072 - mean_squared_error: 0.0503 - acc: 0.9450 - val_loss: 0.3195 - val_mean_squared_error: 0.0951 - val_acc: 0.8650\n",
      "Epoch 669/3000\n",
      " - 0s - loss: 0.2003 - mean_squared_error: 0.0481 - acc: 0.9499 - val_loss: 0.3194 - val_mean_squared_error: 0.0951 - val_acc: 0.8650\n",
      "Epoch 670/3000\n",
      " - 0s - loss: 0.1998 - mean_squared_error: 0.0479 - acc: 0.9450 - val_loss: 0.3192 - val_mean_squared_error: 0.0950 - val_acc: 0.8650\n",
      "Epoch 671/3000\n",
      " - 0s - loss: 0.2028 - mean_squared_error: 0.0484 - acc: 0.9523 - val_loss: 0.3191 - val_mean_squared_error: 0.0950 - val_acc: 0.8650\n",
      "Epoch 672/3000\n",
      " - 0s - loss: 0.2031 - mean_squared_error: 0.0489 - acc: 0.9487 - val_loss: 0.3189 - val_mean_squared_error: 0.0950 - val_acc: 0.8650\n",
      "Epoch 673/3000\n",
      " - 0s - loss: 0.1993 - mean_squared_error: 0.0478 - acc: 0.9474 - val_loss: 0.3188 - val_mean_squared_error: 0.0949 - val_acc: 0.8650\n",
      "Epoch 674/3000\n",
      " - 0s - loss: 0.1999 - mean_squared_error: 0.0477 - acc: 0.9462 - val_loss: 0.3187 - val_mean_squared_error: 0.0949 - val_acc: 0.8650\n",
      "Epoch 675/3000\n",
      " - 0s - loss: 0.2036 - mean_squared_error: 0.0493 - acc: 0.9474 - val_loss: 0.3186 - val_mean_squared_error: 0.0949 - val_acc: 0.8650\n",
      "Epoch 676/3000\n",
      " - 0s - loss: 0.2002 - mean_squared_error: 0.0479 - acc: 0.9535 - val_loss: 0.3185 - val_mean_squared_error: 0.0949 - val_acc: 0.8650\n",
      "Epoch 677/3000\n",
      " - 0s - loss: 0.2022 - mean_squared_error: 0.0483 - acc: 0.9511 - val_loss: 0.3185 - val_mean_squared_error: 0.0949 - val_acc: 0.8650\n",
      "Epoch 678/3000\n",
      " - 0s - loss: 0.1987 - mean_squared_error: 0.0477 - acc: 0.9499 - val_loss: 0.3184 - val_mean_squared_error: 0.0949 - val_acc: 0.8650\n",
      "Epoch 679/3000\n",
      " - 0s - loss: 0.2001 - mean_squared_error: 0.0478 - acc: 0.9523 - val_loss: 0.3184 - val_mean_squared_error: 0.0948 - val_acc: 0.8650\n",
      "Epoch 680/3000\n",
      " - 0s - loss: 0.2007 - mean_squared_error: 0.0485 - acc: 0.9438 - val_loss: 0.3184 - val_mean_squared_error: 0.0948 - val_acc: 0.8650\n",
      "Epoch 681/3000\n",
      " - 0s - loss: 0.2025 - mean_squared_error: 0.0488 - acc: 0.9487 - val_loss: 0.3183 - val_mean_squared_error: 0.0948 - val_acc: 0.8650\n",
      "Epoch 682/3000\n",
      " - 0s - loss: 0.1958 - mean_squared_error: 0.0459 - acc: 0.9609 - val_loss: 0.3183 - val_mean_squared_error: 0.0948 - val_acc: 0.8650\n",
      "Epoch 683/3000\n",
      " - 0s - loss: 0.2003 - mean_squared_error: 0.0479 - acc: 0.9499 - val_loss: 0.3183 - val_mean_squared_error: 0.0948 - val_acc: 0.8650\n",
      "Epoch 684/3000\n",
      " - 0s - loss: 0.2020 - mean_squared_error: 0.0483 - acc: 0.9499 - val_loss: 0.3182 - val_mean_squared_error: 0.0948 - val_acc: 0.8650\n",
      "Epoch 685/3000\n",
      " - 0s - loss: 0.2010 - mean_squared_error: 0.0480 - acc: 0.9511 - val_loss: 0.3181 - val_mean_squared_error: 0.0948 - val_acc: 0.8650\n",
      "Epoch 686/3000\n",
      " - 0s - loss: 0.1990 - mean_squared_error: 0.0481 - acc: 0.9425 - val_loss: 0.3180 - val_mean_squared_error: 0.0948 - val_acc: 0.8650\n",
      "Epoch 687/3000\n",
      " - 0s - loss: 0.2002 - mean_squared_error: 0.0484 - acc: 0.9450 - val_loss: 0.3179 - val_mean_squared_error: 0.0948 - val_acc: 0.8650\n",
      "Epoch 688/3000\n",
      " - 0s - loss: 0.1978 - mean_squared_error: 0.0470 - acc: 0.9511 - val_loss: 0.3178 - val_mean_squared_error: 0.0947 - val_acc: 0.8650\n",
      "Epoch 689/3000\n",
      " - 0s - loss: 0.1962 - mean_squared_error: 0.0463 - acc: 0.9523 - val_loss: 0.3177 - val_mean_squared_error: 0.0947 - val_acc: 0.8650\n",
      "Epoch 690/3000\n",
      " - 0s - loss: 0.1982 - mean_squared_error: 0.0473 - acc: 0.9535 - val_loss: 0.3176 - val_mean_squared_error: 0.0947 - val_acc: 0.8650\n",
      "Epoch 691/3000\n",
      " - 0s - loss: 0.2034 - mean_squared_error: 0.0497 - acc: 0.9438 - val_loss: 0.3175 - val_mean_squared_error: 0.0947 - val_acc: 0.8650\n",
      "Epoch 692/3000\n",
      " - 0s - loss: 0.2004 - mean_squared_error: 0.0486 - acc: 0.9511 - val_loss: 0.3173 - val_mean_squared_error: 0.0946 - val_acc: 0.8650\n",
      "Epoch 693/3000\n",
      " - 0s - loss: 0.1965 - mean_squared_error: 0.0469 - acc: 0.9438 - val_loss: 0.3172 - val_mean_squared_error: 0.0946 - val_acc: 0.8650\n",
      "Epoch 694/3000\n",
      " - 0s - loss: 0.2006 - mean_squared_error: 0.0482 - acc: 0.9511 - val_loss: 0.3171 - val_mean_squared_error: 0.0946 - val_acc: 0.8650\n",
      "Epoch 695/3000\n",
      " - 0s - loss: 0.1983 - mean_squared_error: 0.0477 - acc: 0.9523 - val_loss: 0.3170 - val_mean_squared_error: 0.0946 - val_acc: 0.8650\n",
      "Epoch 696/3000\n",
      " - 0s - loss: 0.1936 - mean_squared_error: 0.0459 - acc: 0.9572 - val_loss: 0.3170 - val_mean_squared_error: 0.0946 - val_acc: 0.8650\n",
      "Epoch 697/3000\n",
      " - 0s - loss: 0.1981 - mean_squared_error: 0.0470 - acc: 0.9523 - val_loss: 0.3169 - val_mean_squared_error: 0.0946 - val_acc: 0.8650\n",
      "Epoch 698/3000\n",
      " - 0s - loss: 0.2033 - mean_squared_error: 0.0497 - acc: 0.9450 - val_loss: 0.3169 - val_mean_squared_error: 0.0946 - val_acc: 0.8650\n",
      "Epoch 699/3000\n",
      " - 0s - loss: 0.1939 - mean_squared_error: 0.0460 - acc: 0.9523 - val_loss: 0.3168 - val_mean_squared_error: 0.0945 - val_acc: 0.8650\n",
      "Epoch 700/3000\n",
      " - 0s - loss: 0.1924 - mean_squared_error: 0.0456 - acc: 0.9487 - val_loss: 0.3168 - val_mean_squared_error: 0.0945 - val_acc: 0.8650\n",
      "Epoch 701/3000\n",
      " - 0s - loss: 0.1970 - mean_squared_error: 0.0472 - acc: 0.9523 - val_loss: 0.3168 - val_mean_squared_error: 0.0945 - val_acc: 0.8650\n",
      "Epoch 702/3000\n",
      " - 0s - loss: 0.1982 - mean_squared_error: 0.0473 - acc: 0.9487 - val_loss: 0.3168 - val_mean_squared_error: 0.0945 - val_acc: 0.8650\n",
      "Epoch 703/3000\n",
      " - 0s - loss: 0.1948 - mean_squared_error: 0.0466 - acc: 0.9487 - val_loss: 0.3167 - val_mean_squared_error: 0.0945 - val_acc: 0.8650\n",
      "Epoch 704/3000\n",
      " - 0s - loss: 0.1972 - mean_squared_error: 0.0472 - acc: 0.9523 - val_loss: 0.3167 - val_mean_squared_error: 0.0945 - val_acc: 0.8650\n",
      "Epoch 705/3000\n",
      " - 0s - loss: 0.1959 - mean_squared_error: 0.0467 - acc: 0.9535 - val_loss: 0.3166 - val_mean_squared_error: 0.0945 - val_acc: 0.8650\n",
      "Epoch 706/3000\n",
      " - 0s - loss: 0.1965 - mean_squared_error: 0.0468 - acc: 0.9548 - val_loss: 0.3165 - val_mean_squared_error: 0.0945 - val_acc: 0.8650\n",
      "Epoch 707/3000\n",
      " - 0s - loss: 0.1951 - mean_squared_error: 0.0465 - acc: 0.9462 - val_loss: 0.3164 - val_mean_squared_error: 0.0945 - val_acc: 0.8650\n",
      "Epoch 708/3000\n",
      " - 0s - loss: 0.1896 - mean_squared_error: 0.0444 - acc: 0.9548 - val_loss: 0.3163 - val_mean_squared_error: 0.0944 - val_acc: 0.8650\n",
      "Epoch 709/3000\n",
      " - 0s - loss: 0.2008 - mean_squared_error: 0.0489 - acc: 0.9474 - val_loss: 0.3162 - val_mean_squared_error: 0.0944 - val_acc: 0.8650\n",
      "Epoch 710/3000\n",
      " - 0s - loss: 0.1967 - mean_squared_error: 0.0468 - acc: 0.9511 - val_loss: 0.3161 - val_mean_squared_error: 0.0944 - val_acc: 0.8650\n",
      "Epoch 711/3000\n",
      " - 0s - loss: 0.1943 - mean_squared_error: 0.0463 - acc: 0.9560 - val_loss: 0.3159 - val_mean_squared_error: 0.0944 - val_acc: 0.8650\n",
      "Epoch 712/3000\n",
      " - 0s - loss: 0.1964 - mean_squared_error: 0.0468 - acc: 0.9511 - val_loss: 0.3158 - val_mean_squared_error: 0.0944 - val_acc: 0.8650\n",
      "Epoch 713/3000\n",
      " - 0s - loss: 0.1900 - mean_squared_error: 0.0446 - acc: 0.9511 - val_loss: 0.3157 - val_mean_squared_error: 0.0943 - val_acc: 0.8650\n",
      "Epoch 714/3000\n",
      " - 0s - loss: 0.1922 - mean_squared_error: 0.0457 - acc: 0.9499 - val_loss: 0.3156 - val_mean_squared_error: 0.0943 - val_acc: 0.8650\n",
      "Epoch 715/3000\n",
      " - 0s - loss: 0.1933 - mean_squared_error: 0.0458 - acc: 0.9560 - val_loss: 0.3155 - val_mean_squared_error: 0.0943 - val_acc: 0.8650\n",
      "Epoch 716/3000\n",
      " - 0s - loss: 0.1883 - mean_squared_error: 0.0442 - acc: 0.9535 - val_loss: 0.3155 - val_mean_squared_error: 0.0943 - val_acc: 0.8650\n",
      "Epoch 717/3000\n",
      " - 0s - loss: 0.1949 - mean_squared_error: 0.0464 - acc: 0.9487 - val_loss: 0.3154 - val_mean_squared_error: 0.0943 - val_acc: 0.8650\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 718/3000\n",
      " - 0s - loss: 0.1929 - mean_squared_error: 0.0454 - acc: 0.9511 - val_loss: 0.3154 - val_mean_squared_error: 0.0943 - val_acc: 0.8650\n",
      "Epoch 719/3000\n",
      " - 0s - loss: 0.1917 - mean_squared_error: 0.0453 - acc: 0.9535 - val_loss: 0.3153 - val_mean_squared_error: 0.0943 - val_acc: 0.8650\n",
      "Epoch 720/3000\n",
      " - 0s - loss: 0.1924 - mean_squared_error: 0.0456 - acc: 0.9548 - val_loss: 0.3153 - val_mean_squared_error: 0.0943 - val_acc: 0.8650\n",
      "Epoch 721/3000\n",
      " - 0s - loss: 0.1913 - mean_squared_error: 0.0452 - acc: 0.9499 - val_loss: 0.3153 - val_mean_squared_error: 0.0943 - val_acc: 0.8650\n",
      "Epoch 722/3000\n",
      " - 0s - loss: 0.1908 - mean_squared_error: 0.0449 - acc: 0.9572 - val_loss: 0.3153 - val_mean_squared_error: 0.0943 - val_acc: 0.8650\n",
      "Epoch 723/3000\n",
      " - 0s - loss: 0.1922 - mean_squared_error: 0.0455 - acc: 0.9560 - val_loss: 0.3152 - val_mean_squared_error: 0.0943 - val_acc: 0.8613\n",
      "Epoch 724/3000\n",
      " - 0s - loss: 0.1927 - mean_squared_error: 0.0467 - acc: 0.9474 - val_loss: 0.3152 - val_mean_squared_error: 0.0942 - val_acc: 0.8613\n",
      "Epoch 725/3000\n",
      " - 0s - loss: 0.1893 - mean_squared_error: 0.0447 - acc: 0.9499 - val_loss: 0.3151 - val_mean_squared_error: 0.0942 - val_acc: 0.8613\n",
      "Epoch 726/3000\n",
      " - 0s - loss: 0.1944 - mean_squared_error: 0.0469 - acc: 0.9511 - val_loss: 0.3151 - val_mean_squared_error: 0.0942 - val_acc: 0.8613\n",
      "Epoch 727/3000\n",
      " - 0s - loss: 0.1889 - mean_squared_error: 0.0446 - acc: 0.9548 - val_loss: 0.3150 - val_mean_squared_error: 0.0942 - val_acc: 0.8613\n",
      "Epoch 728/3000\n",
      " - 0s - loss: 0.1959 - mean_squared_error: 0.0474 - acc: 0.9438 - val_loss: 0.3149 - val_mean_squared_error: 0.0942 - val_acc: 0.8613\n",
      "Epoch 729/3000\n",
      " - 0s - loss: 0.1940 - mean_squared_error: 0.0463 - acc: 0.9523 - val_loss: 0.3148 - val_mean_squared_error: 0.0942 - val_acc: 0.8613\n",
      "Epoch 730/3000\n",
      " - 0s - loss: 0.1914 - mean_squared_error: 0.0459 - acc: 0.9487 - val_loss: 0.3147 - val_mean_squared_error: 0.0942 - val_acc: 0.8613\n",
      "Epoch 731/3000\n",
      " - 0s - loss: 0.1934 - mean_squared_error: 0.0466 - acc: 0.9438 - val_loss: 0.3146 - val_mean_squared_error: 0.0941 - val_acc: 0.8613\n",
      "Epoch 732/3000\n",
      " - 0s - loss: 0.1937 - mean_squared_error: 0.0469 - acc: 0.9462 - val_loss: 0.3145 - val_mean_squared_error: 0.0941 - val_acc: 0.8613\n",
      "Epoch 733/3000\n",
      " - 0s - loss: 0.1918 - mean_squared_error: 0.0455 - acc: 0.9523 - val_loss: 0.3144 - val_mean_squared_error: 0.0941 - val_acc: 0.8613\n",
      "Epoch 734/3000\n",
      " - 0s - loss: 0.1893 - mean_squared_error: 0.0448 - acc: 0.9572 - val_loss: 0.3143 - val_mean_squared_error: 0.0941 - val_acc: 0.8613\n",
      "Epoch 735/3000\n",
      " - 0s - loss: 0.1870 - mean_squared_error: 0.0437 - acc: 0.9535 - val_loss: 0.3142 - val_mean_squared_error: 0.0941 - val_acc: 0.8613\n",
      "Epoch 736/3000\n",
      " - 0s - loss: 0.1910 - mean_squared_error: 0.0459 - acc: 0.9450 - val_loss: 0.3142 - val_mean_squared_error: 0.0941 - val_acc: 0.8613\n",
      "Epoch 737/3000\n",
      " - 0s - loss: 0.1813 - mean_squared_error: 0.0417 - acc: 0.9572 - val_loss: 0.3141 - val_mean_squared_error: 0.0941 - val_acc: 0.8613\n",
      "Epoch 738/3000\n",
      " - 0s - loss: 0.1863 - mean_squared_error: 0.0434 - acc: 0.9511 - val_loss: 0.3141 - val_mean_squared_error: 0.0941 - val_acc: 0.8613\n",
      "Epoch 739/3000\n",
      " - 0s - loss: 0.1886 - mean_squared_error: 0.0442 - acc: 0.9560 - val_loss: 0.3141 - val_mean_squared_error: 0.0940 - val_acc: 0.8613\n",
      "Epoch 740/3000\n",
      " - 0s - loss: 0.1877 - mean_squared_error: 0.0441 - acc: 0.9499 - val_loss: 0.3140 - val_mean_squared_error: 0.0940 - val_acc: 0.8613\n",
      "Epoch 741/3000\n",
      " - 0s - loss: 0.1878 - mean_squared_error: 0.0445 - acc: 0.9511 - val_loss: 0.3140 - val_mean_squared_error: 0.0940 - val_acc: 0.8613\n",
      "Epoch 742/3000\n",
      " - 0s - loss: 0.1901 - mean_squared_error: 0.0452 - acc: 0.9572 - val_loss: 0.3140 - val_mean_squared_error: 0.0940 - val_acc: 0.8613\n",
      "Epoch 743/3000\n",
      " - 0s - loss: 0.1887 - mean_squared_error: 0.0445 - acc: 0.9535 - val_loss: 0.3140 - val_mean_squared_error: 0.0940 - val_acc: 0.8613\n",
      "Epoch 744/3000\n",
      " - 0s - loss: 0.1876 - mean_squared_error: 0.0447 - acc: 0.9523 - val_loss: 0.3139 - val_mean_squared_error: 0.0940 - val_acc: 0.8613\n",
      "Epoch 745/3000\n",
      " - 0s - loss: 0.1893 - mean_squared_error: 0.0444 - acc: 0.9572 - val_loss: 0.3139 - val_mean_squared_error: 0.0940 - val_acc: 0.8613\n",
      "Epoch 746/3000\n",
      " - 0s - loss: 0.1884 - mean_squared_error: 0.0444 - acc: 0.9511 - val_loss: 0.3138 - val_mean_squared_error: 0.0940 - val_acc: 0.8613\n",
      "Epoch 747/3000\n",
      " - 0s - loss: 0.1865 - mean_squared_error: 0.0439 - acc: 0.9572 - val_loss: 0.3137 - val_mean_squared_error: 0.0940 - val_acc: 0.8613\n",
      "Epoch 748/3000\n",
      " - 0s - loss: 0.1921 - mean_squared_error: 0.0458 - acc: 0.9499 - val_loss: 0.3137 - val_mean_squared_error: 0.0940 - val_acc: 0.8613\n",
      "Epoch 749/3000\n",
      " - 0s - loss: 0.1862 - mean_squared_error: 0.0440 - acc: 0.9499 - val_loss: 0.3136 - val_mean_squared_error: 0.0940 - val_acc: 0.8613\n",
      "Epoch 750/3000\n",
      " - 0s - loss: 0.1817 - mean_squared_error: 0.0428 - acc: 0.9511 - val_loss: 0.3135 - val_mean_squared_error: 0.0940 - val_acc: 0.8613\n",
      "Epoch 751/3000\n",
      " - 0s - loss: 0.1873 - mean_squared_error: 0.0443 - acc: 0.9560 - val_loss: 0.3134 - val_mean_squared_error: 0.0939 - val_acc: 0.8613\n",
      "Epoch 752/3000\n",
      " - 0s - loss: 0.1860 - mean_squared_error: 0.0437 - acc: 0.9523 - val_loss: 0.3133 - val_mean_squared_error: 0.0939 - val_acc: 0.8613\n",
      "Epoch 753/3000\n",
      " - 0s - loss: 0.1894 - mean_squared_error: 0.0452 - acc: 0.9474 - val_loss: 0.3132 - val_mean_squared_error: 0.0939 - val_acc: 0.8613\n",
      "Epoch 754/3000\n",
      " - 0s - loss: 0.1843 - mean_squared_error: 0.0432 - acc: 0.9572 - val_loss: 0.3131 - val_mean_squared_error: 0.0939 - val_acc: 0.8613\n",
      "Epoch 755/3000\n",
      " - 0s - loss: 0.1895 - mean_squared_error: 0.0451 - acc: 0.9523 - val_loss: 0.3130 - val_mean_squared_error: 0.0939 - val_acc: 0.8613\n",
      "Epoch 756/3000\n",
      " - 0s - loss: 0.1853 - mean_squared_error: 0.0437 - acc: 0.9560 - val_loss: 0.3130 - val_mean_squared_error: 0.0939 - val_acc: 0.8613\n",
      "Epoch 757/3000\n",
      " - 0s - loss: 0.1853 - mean_squared_error: 0.0441 - acc: 0.9535 - val_loss: 0.3129 - val_mean_squared_error: 0.0939 - val_acc: 0.8613\n",
      "Epoch 758/3000\n",
      " - 0s - loss: 0.1843 - mean_squared_error: 0.0433 - acc: 0.9511 - val_loss: 0.3129 - val_mean_squared_error: 0.0939 - val_acc: 0.8613\n",
      "Epoch 759/3000\n",
      " - 0s - loss: 0.1845 - mean_squared_error: 0.0437 - acc: 0.9511 - val_loss: 0.3129 - val_mean_squared_error: 0.0939 - val_acc: 0.8613\n",
      "Epoch 760/3000\n",
      " - 0s - loss: 0.1837 - mean_squared_error: 0.0428 - acc: 0.9523 - val_loss: 0.3128 - val_mean_squared_error: 0.0939 - val_acc: 0.8613\n",
      "Epoch 761/3000\n",
      " - 0s - loss: 0.1849 - mean_squared_error: 0.0443 - acc: 0.9548 - val_loss: 0.3128 - val_mean_squared_error: 0.0938 - val_acc: 0.8613\n",
      "Epoch 762/3000\n",
      " - 0s - loss: 0.1836 - mean_squared_error: 0.0429 - acc: 0.9572 - val_loss: 0.3128 - val_mean_squared_error: 0.0938 - val_acc: 0.8613\n",
      "Epoch 763/3000\n",
      " - 0s - loss: 0.1818 - mean_squared_error: 0.0426 - acc: 0.9572 - val_loss: 0.3128 - val_mean_squared_error: 0.0938 - val_acc: 0.8613\n",
      "Epoch 764/3000\n",
      " - 0s - loss: 0.1891 - mean_squared_error: 0.0450 - acc: 0.9511 - val_loss: 0.3127 - val_mean_squared_error: 0.0938 - val_acc: 0.8613\n",
      "Epoch 765/3000\n",
      " - 0s - loss: 0.1837 - mean_squared_error: 0.0431 - acc: 0.9535 - val_loss: 0.3127 - val_mean_squared_error: 0.0938 - val_acc: 0.8613\n",
      "Epoch 766/3000\n",
      " - 0s - loss: 0.1831 - mean_squared_error: 0.0429 - acc: 0.9535 - val_loss: 0.3126 - val_mean_squared_error: 0.0938 - val_acc: 0.8613\n",
      "Epoch 767/3000\n",
      " - 0s - loss: 0.1839 - mean_squared_error: 0.0430 - acc: 0.9609 - val_loss: 0.3126 - val_mean_squared_error: 0.0938 - val_acc: 0.8613\n",
      "Epoch 768/3000\n",
      " - 0s - loss: 0.1845 - mean_squared_error: 0.0434 - acc: 0.9572 - val_loss: 0.3125 - val_mean_squared_error: 0.0938 - val_acc: 0.8613\n",
      "Epoch 769/3000\n",
      " - 0s - loss: 0.1850 - mean_squared_error: 0.0441 - acc: 0.9499 - val_loss: 0.3124 - val_mean_squared_error: 0.0938 - val_acc: 0.8613\n",
      "Epoch 770/3000\n",
      " - 0s - loss: 0.1808 - mean_squared_error: 0.0422 - acc: 0.9548 - val_loss: 0.3123 - val_mean_squared_error: 0.0938 - val_acc: 0.8613\n",
      "Epoch 771/3000\n",
      " - 0s - loss: 0.1855 - mean_squared_error: 0.0443 - acc: 0.9487 - val_loss: 0.3122 - val_mean_squared_error: 0.0938 - val_acc: 0.8613\n",
      "Epoch 772/3000\n",
      " - 0s - loss: 0.1819 - mean_squared_error: 0.0428 - acc: 0.9535 - val_loss: 0.3121 - val_mean_squared_error: 0.0937 - val_acc: 0.8613\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 773/3000\n",
      " - 0s - loss: 0.1804 - mean_squared_error: 0.0421 - acc: 0.9560 - val_loss: 0.3121 - val_mean_squared_error: 0.0937 - val_acc: 0.8613\n",
      "Epoch 774/3000\n",
      " - 0s - loss: 0.1860 - mean_squared_error: 0.0441 - acc: 0.9535 - val_loss: 0.3120 - val_mean_squared_error: 0.0937 - val_acc: 0.8577\n",
      "Epoch 775/3000\n",
      " - 0s - loss: 0.1809 - mean_squared_error: 0.0419 - acc: 0.9535 - val_loss: 0.3119 - val_mean_squared_error: 0.0937 - val_acc: 0.8577\n",
      "Epoch 776/3000\n",
      " - 0s - loss: 0.1820 - mean_squared_error: 0.0432 - acc: 0.9572 - val_loss: 0.3119 - val_mean_squared_error: 0.0937 - val_acc: 0.8577\n",
      "Epoch 777/3000\n",
      " - 0s - loss: 0.1795 - mean_squared_error: 0.0417 - acc: 0.9548 - val_loss: 0.3118 - val_mean_squared_error: 0.0937 - val_acc: 0.8577\n",
      "Epoch 778/3000\n",
      " - 0s - loss: 0.1779 - mean_squared_error: 0.0416 - acc: 0.9535 - val_loss: 0.3118 - val_mean_squared_error: 0.0937 - val_acc: 0.8577\n",
      "Epoch 779/3000\n",
      " - 0s - loss: 0.1805 - mean_squared_error: 0.0425 - acc: 0.9548 - val_loss: 0.3118 - val_mean_squared_error: 0.0937 - val_acc: 0.8577\n",
      "Epoch 780/3000\n",
      " - 0s - loss: 0.1808 - mean_squared_error: 0.0424 - acc: 0.9499 - val_loss: 0.3118 - val_mean_squared_error: 0.0937 - val_acc: 0.8577\n",
      "Epoch 781/3000\n",
      " - 0s - loss: 0.1791 - mean_squared_error: 0.0423 - acc: 0.9487 - val_loss: 0.3118 - val_mean_squared_error: 0.0937 - val_acc: 0.8577\n",
      "Epoch 782/3000\n",
      " - 0s - loss: 0.1796 - mean_squared_error: 0.0422 - acc: 0.9523 - val_loss: 0.3117 - val_mean_squared_error: 0.0937 - val_acc: 0.8577\n",
      "Epoch 783/3000\n",
      " - 0s - loss: 0.1784 - mean_squared_error: 0.0416 - acc: 0.9560 - val_loss: 0.3117 - val_mean_squared_error: 0.0937 - val_acc: 0.8577\n",
      "Epoch 784/3000\n",
      " - 0s - loss: 0.1835 - mean_squared_error: 0.0429 - acc: 0.9499 - val_loss: 0.3117 - val_mean_squared_error: 0.0937 - val_acc: 0.8577\n",
      "Epoch 785/3000\n",
      " - 0s - loss: 0.1780 - mean_squared_error: 0.0417 - acc: 0.9535 - val_loss: 0.3116 - val_mean_squared_error: 0.0937 - val_acc: 0.8577\n",
      "Epoch 786/3000\n",
      " - 0s - loss: 0.1781 - mean_squared_error: 0.0414 - acc: 0.9560 - val_loss: 0.3116 - val_mean_squared_error: 0.0937 - val_acc: 0.8577\n",
      "Epoch 787/3000\n",
      " - 0s - loss: 0.1808 - mean_squared_error: 0.0425 - acc: 0.9511 - val_loss: 0.3115 - val_mean_squared_error: 0.0936 - val_acc: 0.8577\n",
      "Epoch 788/3000\n",
      " - 0s - loss: 0.1783 - mean_squared_error: 0.0417 - acc: 0.9548 - val_loss: 0.3114 - val_mean_squared_error: 0.0936 - val_acc: 0.8577\n",
      "Epoch 789/3000\n",
      " - 0s - loss: 0.1803 - mean_squared_error: 0.0420 - acc: 0.9560 - val_loss: 0.3113 - val_mean_squared_error: 0.0936 - val_acc: 0.8577\n",
      "Epoch 790/3000\n",
      " - 0s - loss: 0.1785 - mean_squared_error: 0.0411 - acc: 0.9584 - val_loss: 0.3113 - val_mean_squared_error: 0.0936 - val_acc: 0.8577\n",
      "Epoch 791/3000\n",
      " - 0s - loss: 0.1826 - mean_squared_error: 0.0431 - acc: 0.9560 - val_loss: 0.3112 - val_mean_squared_error: 0.0936 - val_acc: 0.8577\n",
      "Epoch 792/3000\n",
      " - 0s - loss: 0.1750 - mean_squared_error: 0.0403 - acc: 0.9572 - val_loss: 0.3111 - val_mean_squared_error: 0.0936 - val_acc: 0.8577\n",
      "Epoch 793/3000\n",
      " - 0s - loss: 0.1763 - mean_squared_error: 0.0410 - acc: 0.9572 - val_loss: 0.3110 - val_mean_squared_error: 0.0936 - val_acc: 0.8577\n",
      "Epoch 794/3000\n",
      " - 0s - loss: 0.1805 - mean_squared_error: 0.0425 - acc: 0.9535 - val_loss: 0.3109 - val_mean_squared_error: 0.0936 - val_acc: 0.8577\n",
      "Epoch 795/3000\n",
      " - 0s - loss: 0.1797 - mean_squared_error: 0.0419 - acc: 0.9560 - val_loss: 0.3109 - val_mean_squared_error: 0.0936 - val_acc: 0.8577\n",
      "Epoch 796/3000\n",
      " - 0s - loss: 0.1754 - mean_squared_error: 0.0410 - acc: 0.9597 - val_loss: 0.3108 - val_mean_squared_error: 0.0935 - val_acc: 0.8577\n",
      "Epoch 797/3000\n",
      " - 0s - loss: 0.1792 - mean_squared_error: 0.0421 - acc: 0.9535 - val_loss: 0.3108 - val_mean_squared_error: 0.0935 - val_acc: 0.8577\n",
      "Epoch 798/3000\n",
      " - 0s - loss: 0.1836 - mean_squared_error: 0.0439 - acc: 0.9462 - val_loss: 0.3108 - val_mean_squared_error: 0.0935 - val_acc: 0.8577\n",
      "Epoch 799/3000\n",
      " - 0s - loss: 0.1753 - mean_squared_error: 0.0406 - acc: 0.9621 - val_loss: 0.3107 - val_mean_squared_error: 0.0935 - val_acc: 0.8577\n",
      "Epoch 800/3000\n",
      " - 0s - loss: 0.1744 - mean_squared_error: 0.0403 - acc: 0.9584 - val_loss: 0.3107 - val_mean_squared_error: 0.0935 - val_acc: 0.8577\n",
      "Epoch 801/3000\n",
      " - 0s - loss: 0.1726 - mean_squared_error: 0.0392 - acc: 0.9572 - val_loss: 0.3107 - val_mean_squared_error: 0.0935 - val_acc: 0.8577\n",
      "Epoch 802/3000\n",
      " - 0s - loss: 0.1794 - mean_squared_error: 0.0424 - acc: 0.9499 - val_loss: 0.3107 - val_mean_squared_error: 0.0935 - val_acc: 0.8577\n",
      "Epoch 803/3000\n",
      " - 0s - loss: 0.1786 - mean_squared_error: 0.0411 - acc: 0.9645 - val_loss: 0.3107 - val_mean_squared_error: 0.0935 - val_acc: 0.8577\n",
      "Epoch 804/3000\n",
      " - 0s - loss: 0.1771 - mean_squared_error: 0.0416 - acc: 0.9487 - val_loss: 0.3106 - val_mean_squared_error: 0.0935 - val_acc: 0.8577\n",
      "Epoch 805/3000\n",
      " - 0s - loss: 0.1698 - mean_squared_error: 0.0388 - acc: 0.9621 - val_loss: 0.3106 - val_mean_squared_error: 0.0935 - val_acc: 0.8577\n",
      "Epoch 806/3000\n",
      " - 0s - loss: 0.1707 - mean_squared_error: 0.0395 - acc: 0.9560 - val_loss: 0.3105 - val_mean_squared_error: 0.0935 - val_acc: 0.8577\n",
      "Epoch 807/3000\n",
      " - 0s - loss: 0.1748 - mean_squared_error: 0.0407 - acc: 0.9560 - val_loss: 0.3105 - val_mean_squared_error: 0.0935 - val_acc: 0.8577\n",
      "Epoch 808/3000\n",
      " - 0s - loss: 0.1772 - mean_squared_error: 0.0416 - acc: 0.9548 - val_loss: 0.3104 - val_mean_squared_error: 0.0935 - val_acc: 0.8577\n",
      "Epoch 809/3000\n",
      " - 0s - loss: 0.1735 - mean_squared_error: 0.0406 - acc: 0.9523 - val_loss: 0.3103 - val_mean_squared_error: 0.0935 - val_acc: 0.8577\n",
      "Epoch 810/3000\n",
      " - 0s - loss: 0.1720 - mean_squared_error: 0.0395 - acc: 0.9584 - val_loss: 0.3103 - val_mean_squared_error: 0.0935 - val_acc: 0.8577\n",
      "Epoch 811/3000\n",
      " - 0s - loss: 0.1823 - mean_squared_error: 0.0440 - acc: 0.9474 - val_loss: 0.3102 - val_mean_squared_error: 0.0935 - val_acc: 0.8577\n",
      "Epoch 812/3000\n",
      " - 0s - loss: 0.1795 - mean_squared_error: 0.0425 - acc: 0.9584 - val_loss: 0.3101 - val_mean_squared_error: 0.0934 - val_acc: 0.8577\n",
      "Epoch 813/3000\n",
      " - 0s - loss: 0.1759 - mean_squared_error: 0.0409 - acc: 0.9535 - val_loss: 0.3100 - val_mean_squared_error: 0.0934 - val_acc: 0.8577\n",
      "Epoch 814/3000\n",
      " - 0s - loss: 0.1735 - mean_squared_error: 0.0393 - acc: 0.9633 - val_loss: 0.3100 - val_mean_squared_error: 0.0934 - val_acc: 0.8577\n",
      "Epoch 815/3000\n",
      " - 0s - loss: 0.1754 - mean_squared_error: 0.0412 - acc: 0.9548 - val_loss: 0.3099 - val_mean_squared_error: 0.0934 - val_acc: 0.8577\n",
      "Epoch 816/3000\n",
      " - 0s - loss: 0.1724 - mean_squared_error: 0.0399 - acc: 0.9584 - val_loss: 0.3099 - val_mean_squared_error: 0.0934 - val_acc: 0.8577\n",
      "Epoch 817/3000\n",
      " - 0s - loss: 0.1795 - mean_squared_error: 0.0427 - acc: 0.9572 - val_loss: 0.3098 - val_mean_squared_error: 0.0934 - val_acc: 0.8577\n",
      "Epoch 818/3000\n",
      " - 0s - loss: 0.1737 - mean_squared_error: 0.0402 - acc: 0.9597 - val_loss: 0.3098 - val_mean_squared_error: 0.0934 - val_acc: 0.8577\n",
      "Epoch 819/3000\n",
      " - 0s - loss: 0.1750 - mean_squared_error: 0.0409 - acc: 0.9560 - val_loss: 0.3098 - val_mean_squared_error: 0.0934 - val_acc: 0.8577\n",
      "Epoch 820/3000\n",
      " - 0s - loss: 0.1704 - mean_squared_error: 0.0389 - acc: 0.9609 - val_loss: 0.3098 - val_mean_squared_error: 0.0934 - val_acc: 0.8577\n",
      "Epoch 821/3000\n",
      " - 0s - loss: 0.1751 - mean_squared_error: 0.0410 - acc: 0.9584 - val_loss: 0.3098 - val_mean_squared_error: 0.0934 - val_acc: 0.8577\n",
      "Epoch 822/3000\n",
      " - 0s - loss: 0.1797 - mean_squared_error: 0.0423 - acc: 0.9535 - val_loss: 0.3098 - val_mean_squared_error: 0.0934 - val_acc: 0.8577\n",
      "Epoch 823/3000\n",
      " - 0s - loss: 0.1726 - mean_squared_error: 0.0396 - acc: 0.9548 - val_loss: 0.3097 - val_mean_squared_error: 0.0934 - val_acc: 0.8577\n",
      "Epoch 824/3000\n",
      " - 0s - loss: 0.1718 - mean_squared_error: 0.0395 - acc: 0.9609 - val_loss: 0.3097 - val_mean_squared_error: 0.0934 - val_acc: 0.8577\n",
      "Epoch 825/3000\n",
      " - 0s - loss: 0.1743 - mean_squared_error: 0.0403 - acc: 0.9609 - val_loss: 0.3097 - val_mean_squared_error: 0.0934 - val_acc: 0.8577\n",
      "Epoch 826/3000\n",
      " - 0s - loss: 0.1688 - mean_squared_error: 0.0383 - acc: 0.9645 - val_loss: 0.3096 - val_mean_squared_error: 0.0934 - val_acc: 0.8577\n",
      "Epoch 827/3000\n",
      " - 0s - loss: 0.1722 - mean_squared_error: 0.0397 - acc: 0.9597 - val_loss: 0.3096 - val_mean_squared_error: 0.0934 - val_acc: 0.8577\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 828/3000\n",
      " - 0s - loss: 0.1777 - mean_squared_error: 0.0420 - acc: 0.9523 - val_loss: 0.3095 - val_mean_squared_error: 0.0934 - val_acc: 0.8577\n",
      "Epoch 829/3000\n",
      " - 0s - loss: 0.1756 - mean_squared_error: 0.0411 - acc: 0.9560 - val_loss: 0.3095 - val_mean_squared_error: 0.0934 - val_acc: 0.8577\n",
      "Epoch 830/3000\n",
      " - 0s - loss: 0.1730 - mean_squared_error: 0.0401 - acc: 0.9597 - val_loss: 0.3094 - val_mean_squared_error: 0.0934 - val_acc: 0.8577\n",
      "Epoch 831/3000\n",
      " - 0s - loss: 0.1712 - mean_squared_error: 0.0395 - acc: 0.9609 - val_loss: 0.3093 - val_mean_squared_error: 0.0934 - val_acc: 0.8577\n",
      "Epoch 832/3000\n",
      " - 0s - loss: 0.1709 - mean_squared_error: 0.0392 - acc: 0.9645 - val_loss: 0.3093 - val_mean_squared_error: 0.0933 - val_acc: 0.8577\n",
      "Epoch 833/3000\n",
      " - 0s - loss: 0.1701 - mean_squared_error: 0.0394 - acc: 0.9584 - val_loss: 0.3092 - val_mean_squared_error: 0.0933 - val_acc: 0.8577\n",
      "Epoch 834/3000\n",
      " - 0s - loss: 0.1741 - mean_squared_error: 0.0408 - acc: 0.9535 - val_loss: 0.3091 - val_mean_squared_error: 0.0933 - val_acc: 0.8577\n",
      "Epoch 835/3000\n",
      " - 0s - loss: 0.1739 - mean_squared_error: 0.0402 - acc: 0.9584 - val_loss: 0.3091 - val_mean_squared_error: 0.0933 - val_acc: 0.8577\n",
      "Epoch 836/3000\n",
      " - 0s - loss: 0.1722 - mean_squared_error: 0.0396 - acc: 0.9609 - val_loss: 0.3091 - val_mean_squared_error: 0.0933 - val_acc: 0.8577\n",
      "Epoch 837/3000\n",
      " - 0s - loss: 0.1703 - mean_squared_error: 0.0395 - acc: 0.9560 - val_loss: 0.3090 - val_mean_squared_error: 0.0933 - val_acc: 0.8577\n",
      "Epoch 838/3000\n",
      " - 0s - loss: 0.1749 - mean_squared_error: 0.0414 - acc: 0.9609 - val_loss: 0.3090 - val_mean_squared_error: 0.0933 - val_acc: 0.8577\n",
      "Epoch 839/3000\n",
      " - 0s - loss: 0.1695 - mean_squared_error: 0.0394 - acc: 0.9548 - val_loss: 0.3090 - val_mean_squared_error: 0.0933 - val_acc: 0.8577\n",
      "Epoch 840/3000\n",
      " - 0s - loss: 0.1671 - mean_squared_error: 0.0382 - acc: 0.9609 - val_loss: 0.3090 - val_mean_squared_error: 0.0933 - val_acc: 0.8577\n",
      "Epoch 841/3000\n",
      " - 0s - loss: 0.1703 - mean_squared_error: 0.0397 - acc: 0.9572 - val_loss: 0.3089 - val_mean_squared_error: 0.0933 - val_acc: 0.8577\n",
      "Epoch 842/3000\n",
      " - 0s - loss: 0.1730 - mean_squared_error: 0.0409 - acc: 0.9597 - val_loss: 0.3089 - val_mean_squared_error: 0.0933 - val_acc: 0.8577\n",
      "Epoch 843/3000\n",
      " - 0s - loss: 0.1676 - mean_squared_error: 0.0382 - acc: 0.9609 - val_loss: 0.3089 - val_mean_squared_error: 0.0933 - val_acc: 0.8577\n",
      "Epoch 844/3000\n",
      " - 0s - loss: 0.1728 - mean_squared_error: 0.0399 - acc: 0.9584 - val_loss: 0.3089 - val_mean_squared_error: 0.0933 - val_acc: 0.8577\n",
      "Epoch 845/3000\n",
      " - 0s - loss: 0.1644 - mean_squared_error: 0.0373 - acc: 0.9609 - val_loss: 0.3088 - val_mean_squared_error: 0.0933 - val_acc: 0.8577\n",
      "Epoch 846/3000\n",
      " - 0s - loss: 0.1678 - mean_squared_error: 0.0387 - acc: 0.9609 - val_loss: 0.3088 - val_mean_squared_error: 0.0933 - val_acc: 0.8577\n",
      "Epoch 847/3000\n",
      " - 0s - loss: 0.1685 - mean_squared_error: 0.0388 - acc: 0.9597 - val_loss: 0.3087 - val_mean_squared_error: 0.0933 - val_acc: 0.8577\n",
      "Epoch 848/3000\n",
      " - 0s - loss: 0.1658 - mean_squared_error: 0.0377 - acc: 0.9584 - val_loss: 0.3087 - val_mean_squared_error: 0.0933 - val_acc: 0.8577\n",
      "Epoch 849/3000\n",
      " - 0s - loss: 0.1660 - mean_squared_error: 0.0380 - acc: 0.9584 - val_loss: 0.3086 - val_mean_squared_error: 0.0933 - val_acc: 0.8577\n",
      "Epoch 850/3000\n",
      " - 0s - loss: 0.1687 - mean_squared_error: 0.0389 - acc: 0.9597 - val_loss: 0.3085 - val_mean_squared_error: 0.0932 - val_acc: 0.8577\n",
      "Epoch 851/3000\n",
      " - 0s - loss: 0.1681 - mean_squared_error: 0.0391 - acc: 0.9584 - val_loss: 0.3085 - val_mean_squared_error: 0.0932 - val_acc: 0.8577\n",
      "Epoch 852/3000\n",
      " - 0s - loss: 0.1685 - mean_squared_error: 0.0388 - acc: 0.9597 - val_loss: 0.3084 - val_mean_squared_error: 0.0932 - val_acc: 0.8613\n",
      "Epoch 853/3000\n",
      " - 0s - loss: 0.1656 - mean_squared_error: 0.0378 - acc: 0.9597 - val_loss: 0.3083 - val_mean_squared_error: 0.0932 - val_acc: 0.8613\n",
      "Epoch 854/3000\n",
      " - 0s - loss: 0.1647 - mean_squared_error: 0.0374 - acc: 0.9645 - val_loss: 0.3083 - val_mean_squared_error: 0.0932 - val_acc: 0.8613\n",
      "Epoch 855/3000\n",
      " - 0s - loss: 0.1707 - mean_squared_error: 0.0396 - acc: 0.9548 - val_loss: 0.3082 - val_mean_squared_error: 0.0932 - val_acc: 0.8613\n",
      "Epoch 856/3000\n",
      " - 0s - loss: 0.1693 - mean_squared_error: 0.0393 - acc: 0.9548 - val_loss: 0.3082 - val_mean_squared_error: 0.0932 - val_acc: 0.8613\n",
      "Epoch 857/3000\n",
      " - 0s - loss: 0.1664 - mean_squared_error: 0.0381 - acc: 0.9572 - val_loss: 0.3082 - val_mean_squared_error: 0.0932 - val_acc: 0.8613\n",
      "Epoch 858/3000\n",
      " - 0s - loss: 0.1644 - mean_squared_error: 0.0375 - acc: 0.9584 - val_loss: 0.3082 - val_mean_squared_error: 0.0932 - val_acc: 0.8613\n",
      "Epoch 859/3000\n",
      " - 0s - loss: 0.1691 - mean_squared_error: 0.0393 - acc: 0.9572 - val_loss: 0.3081 - val_mean_squared_error: 0.0932 - val_acc: 0.8613\n",
      "Epoch 860/3000\n",
      " - 0s - loss: 0.1641 - mean_squared_error: 0.0370 - acc: 0.9633 - val_loss: 0.3081 - val_mean_squared_error: 0.0932 - val_acc: 0.8613\n",
      "Epoch 861/3000\n",
      " - 0s - loss: 0.1659 - mean_squared_error: 0.0375 - acc: 0.9597 - val_loss: 0.3081 - val_mean_squared_error: 0.0932 - val_acc: 0.8613\n",
      "Epoch 862/3000\n",
      " - 0s - loss: 0.1649 - mean_squared_error: 0.0377 - acc: 0.9572 - val_loss: 0.3081 - val_mean_squared_error: 0.0932 - val_acc: 0.8613\n",
      "Epoch 863/3000\n",
      " - 0s - loss: 0.1659 - mean_squared_error: 0.0384 - acc: 0.9584 - val_loss: 0.3081 - val_mean_squared_error: 0.0932 - val_acc: 0.8613\n",
      "Epoch 864/3000\n",
      " - 0s - loss: 0.1628 - mean_squared_error: 0.0373 - acc: 0.9584 - val_loss: 0.3081 - val_mean_squared_error: 0.0932 - val_acc: 0.8613\n",
      "Epoch 865/3000\n",
      " - 0s - loss: 0.1670 - mean_squared_error: 0.0383 - acc: 0.9572 - val_loss: 0.3080 - val_mean_squared_error: 0.0932 - val_acc: 0.8613\n",
      "Epoch 866/3000\n",
      " - 0s - loss: 0.1645 - mean_squared_error: 0.0372 - acc: 0.9597 - val_loss: 0.3080 - val_mean_squared_error: 0.0932 - val_acc: 0.8613\n",
      "Epoch 867/3000\n",
      " - 0s - loss: 0.1676 - mean_squared_error: 0.0391 - acc: 0.9535 - val_loss: 0.3079 - val_mean_squared_error: 0.0932 - val_acc: 0.8613\n",
      "Epoch 868/3000\n",
      " - 0s - loss: 0.1688 - mean_squared_error: 0.0395 - acc: 0.9523 - val_loss: 0.3079 - val_mean_squared_error: 0.0932 - val_acc: 0.8613\n",
      "Epoch 869/3000\n",
      " - 0s - loss: 0.1642 - mean_squared_error: 0.0371 - acc: 0.9621 - val_loss: 0.3078 - val_mean_squared_error: 0.0932 - val_acc: 0.8613\n",
      "Epoch 870/3000\n",
      " - 0s - loss: 0.1668 - mean_squared_error: 0.0383 - acc: 0.9597 - val_loss: 0.3078 - val_mean_squared_error: 0.0932 - val_acc: 0.8613\n",
      "Epoch 871/3000\n",
      " - 0s - loss: 0.1654 - mean_squared_error: 0.0379 - acc: 0.9609 - val_loss: 0.3077 - val_mean_squared_error: 0.0932 - val_acc: 0.8613\n",
      "Epoch 872/3000\n",
      " - 0s - loss: 0.1667 - mean_squared_error: 0.0387 - acc: 0.9548 - val_loss: 0.3077 - val_mean_squared_error: 0.0932 - val_acc: 0.8613\n",
      "Epoch 873/3000\n",
      " - 0s - loss: 0.1647 - mean_squared_error: 0.0383 - acc: 0.9523 - val_loss: 0.3076 - val_mean_squared_error: 0.0932 - val_acc: 0.8613\n",
      "Epoch 874/3000\n",
      " - 0s - loss: 0.1595 - mean_squared_error: 0.0361 - acc: 0.9597 - val_loss: 0.3076 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 875/3000\n",
      " - 0s - loss: 0.1646 - mean_squared_error: 0.0381 - acc: 0.9633 - val_loss: 0.3075 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 876/3000\n",
      " - 0s - loss: 0.1601 - mean_squared_error: 0.0355 - acc: 0.9682 - val_loss: 0.3075 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 877/3000\n",
      " - 0s - loss: 0.1622 - mean_squared_error: 0.0370 - acc: 0.9621 - val_loss: 0.3075 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 878/3000\n",
      " - 0s - loss: 0.1632 - mean_squared_error: 0.0376 - acc: 0.9584 - val_loss: 0.3075 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 879/3000\n",
      " - 0s - loss: 0.1596 - mean_squared_error: 0.0360 - acc: 0.9621 - val_loss: 0.3075 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 880/3000\n",
      " - 0s - loss: 0.1683 - mean_squared_error: 0.0391 - acc: 0.9584 - val_loss: 0.3074 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 881/3000\n",
      " - 0s - loss: 0.1642 - mean_squared_error: 0.0379 - acc: 0.9597 - val_loss: 0.3074 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 882/3000\n",
      " - 0s - loss: 0.1628 - mean_squared_error: 0.0374 - acc: 0.9572 - val_loss: 0.3074 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 883/3000\n",
      " - 0s - loss: 0.1638 - mean_squared_error: 0.0372 - acc: 0.9609 - val_loss: 0.3074 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 884/3000\n",
      " - 0s - loss: 0.1597 - mean_squared_error: 0.0361 - acc: 0.9597 - val_loss: 0.3074 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 885/3000\n",
      " - 0s - loss: 0.1627 - mean_squared_error: 0.0367 - acc: 0.9560 - val_loss: 0.3074 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 886/3000\n",
      " - 0s - loss: 0.1599 - mean_squared_error: 0.0359 - acc: 0.9621 - val_loss: 0.3073 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 887/3000\n",
      " - 0s - loss: 0.1609 - mean_squared_error: 0.0363 - acc: 0.9597 - val_loss: 0.3073 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 888/3000\n",
      " - 0s - loss: 0.1588 - mean_squared_error: 0.0357 - acc: 0.9645 - val_loss: 0.3072 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 889/3000\n",
      " - 0s - loss: 0.1611 - mean_squared_error: 0.0370 - acc: 0.9584 - val_loss: 0.3072 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 890/3000\n",
      " - 0s - loss: 0.1639 - mean_squared_error: 0.0378 - acc: 0.9584 - val_loss: 0.3071 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 891/3000\n",
      " - 0s - loss: 0.1628 - mean_squared_error: 0.0373 - acc: 0.9597 - val_loss: 0.3071 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 892/3000\n",
      " - 0s - loss: 0.1600 - mean_squared_error: 0.0362 - acc: 0.9633 - val_loss: 0.3070 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 893/3000\n",
      " - 0s - loss: 0.1590 - mean_squared_error: 0.0356 - acc: 0.9609 - val_loss: 0.3070 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 894/3000\n",
      " - 0s - loss: 0.1622 - mean_squared_error: 0.0373 - acc: 0.9584 - val_loss: 0.3070 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 895/3000\n",
      " - 0s - loss: 0.1631 - mean_squared_error: 0.0377 - acc: 0.9572 - val_loss: 0.3069 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 896/3000\n",
      " - 0s - loss: 0.1602 - mean_squared_error: 0.0361 - acc: 0.9633 - val_loss: 0.3069 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 897/3000\n",
      " - 0s - loss: 0.1606 - mean_squared_error: 0.0364 - acc: 0.9658 - val_loss: 0.3069 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 898/3000\n",
      " - 0s - loss: 0.1631 - mean_squared_error: 0.0378 - acc: 0.9597 - val_loss: 0.3068 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 899/3000\n",
      " - 0s - loss: 0.1579 - mean_squared_error: 0.0352 - acc: 0.9658 - val_loss: 0.3068 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 900/3000\n",
      " - 0s - loss: 0.1606 - mean_squared_error: 0.0365 - acc: 0.9560 - val_loss: 0.3068 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 901/3000\n",
      " - 0s - loss: 0.1622 - mean_squared_error: 0.0369 - acc: 0.9609 - val_loss: 0.3068 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 902/3000\n",
      " - 0s - loss: 0.1589 - mean_squared_error: 0.0357 - acc: 0.9633 - val_loss: 0.3068 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 903/3000\n",
      " - 0s - loss: 0.1617 - mean_squared_error: 0.0369 - acc: 0.9609 - val_loss: 0.3068 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 904/3000\n",
      " - 0s - loss: 0.1605 - mean_squared_error: 0.0365 - acc: 0.9597 - val_loss: 0.3068 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 905/3000\n",
      " - 0s - loss: 0.1603 - mean_squared_error: 0.0367 - acc: 0.9572 - val_loss: 0.3067 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 906/3000\n",
      " - 0s - loss: 0.1590 - mean_squared_error: 0.0358 - acc: 0.9645 - val_loss: 0.3067 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 907/3000\n",
      " - 0s - loss: 0.1590 - mean_squared_error: 0.0360 - acc: 0.9597 - val_loss: 0.3067 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 908/3000\n",
      " - 0s - loss: 0.1559 - mean_squared_error: 0.0348 - acc: 0.9633 - val_loss: 0.3066 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 909/3000\n",
      " - 0s - loss: 0.1578 - mean_squared_error: 0.0359 - acc: 0.9609 - val_loss: 0.3066 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 910/3000\n",
      " - 0s - loss: 0.1573 - mean_squared_error: 0.0357 - acc: 0.9584 - val_loss: 0.3066 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 911/3000\n",
      " - 0s - loss: 0.1592 - mean_squared_error: 0.0366 - acc: 0.9584 - val_loss: 0.3065 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 912/3000\n",
      " - 0s - loss: 0.1608 - mean_squared_error: 0.0369 - acc: 0.9621 - val_loss: 0.3065 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 913/3000\n",
      " - 0s - loss: 0.1574 - mean_squared_error: 0.0354 - acc: 0.9609 - val_loss: 0.3064 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 914/3000\n",
      " - 0s - loss: 0.1549 - mean_squared_error: 0.0348 - acc: 0.9633 - val_loss: 0.3064 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 915/3000\n",
      " - 0s - loss: 0.1543 - mean_squared_error: 0.0346 - acc: 0.9621 - val_loss: 0.3064 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 916/3000\n",
      " - 0s - loss: 0.1549 - mean_squared_error: 0.0352 - acc: 0.9670 - val_loss: 0.3063 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 917/3000\n",
      " - 0s - loss: 0.1580 - mean_squared_error: 0.0357 - acc: 0.9621 - val_loss: 0.3063 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 918/3000\n",
      " - 0s - loss: 0.1582 - mean_squared_error: 0.0363 - acc: 0.9645 - val_loss: 0.3063 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 919/3000\n",
      " - 0s - loss: 0.1534 - mean_squared_error: 0.0339 - acc: 0.9707 - val_loss: 0.3063 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 920/3000\n",
      " - 0s - loss: 0.1571 - mean_squared_error: 0.0353 - acc: 0.9645 - val_loss: 0.3063 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 921/3000\n",
      " - 0s - loss: 0.1593 - mean_squared_error: 0.0365 - acc: 0.9621 - val_loss: 0.3063 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 922/3000\n",
      " - 0s - loss: 0.1580 - mean_squared_error: 0.0360 - acc: 0.9670 - val_loss: 0.3063 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 923/3000\n",
      " - 0s - loss: 0.1607 - mean_squared_error: 0.0367 - acc: 0.9584 - val_loss: 0.3062 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 924/3000\n",
      " - 0s - loss: 0.1592 - mean_squared_error: 0.0359 - acc: 0.9609 - val_loss: 0.3062 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 925/3000\n",
      " - 0s - loss: 0.1526 - mean_squared_error: 0.0340 - acc: 0.9658 - val_loss: 0.3062 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 926/3000\n",
      " - 0s - loss: 0.1577 - mean_squared_error: 0.0353 - acc: 0.9633 - val_loss: 0.3062 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 927/3000\n",
      " - 0s - loss: 0.1593 - mean_squared_error: 0.0366 - acc: 0.9572 - val_loss: 0.3062 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 928/3000\n",
      " - 0s - loss: 0.1521 - mean_squared_error: 0.0337 - acc: 0.9682 - val_loss: 0.3061 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 929/3000\n",
      " - 0s - loss: 0.1600 - mean_squared_error: 0.0371 - acc: 0.9560 - val_loss: 0.3061 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 930/3000\n",
      " - 0s - loss: 0.1588 - mean_squared_error: 0.0369 - acc: 0.9597 - val_loss: 0.3060 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 931/3000\n",
      " - 0s - loss: 0.1555 - mean_squared_error: 0.0351 - acc: 0.9645 - val_loss: 0.3060 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 932/3000\n",
      " - 0s - loss: 0.1549 - mean_squared_error: 0.0353 - acc: 0.9633 - val_loss: 0.3060 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 933/3000\n",
      " - 0s - loss: 0.1565 - mean_squared_error: 0.0355 - acc: 0.9572 - val_loss: 0.3059 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 934/3000\n",
      " - 0s - loss: 0.1528 - mean_squared_error: 0.0342 - acc: 0.9633 - val_loss: 0.3059 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 935/3000\n",
      " - 0s - loss: 0.1572 - mean_squared_error: 0.0366 - acc: 0.9548 - val_loss: 0.3059 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 936/3000\n",
      " - 0s - loss: 0.1525 - mean_squared_error: 0.0338 - acc: 0.9609 - val_loss: 0.3058 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 937/3000\n",
      " - 0s - loss: 0.1498 - mean_squared_error: 0.0332 - acc: 0.9670 - val_loss: 0.3058 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 938/3000\n",
      " - 0s - loss: 0.1536 - mean_squared_error: 0.0348 - acc: 0.9621 - val_loss: 0.3058 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 939/3000\n",
      " - 0s - loss: 0.1550 - mean_squared_error: 0.0349 - acc: 0.9670 - val_loss: 0.3058 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 940/3000\n",
      " - 0s - loss: 0.1572 - mean_squared_error: 0.0357 - acc: 0.9609 - val_loss: 0.3058 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 941/3000\n",
      " - 0s - loss: 0.1526 - mean_squared_error: 0.0340 - acc: 0.9609 - val_loss: 0.3058 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 942/3000\n",
      " - 0s - loss: 0.1595 - mean_squared_error: 0.0368 - acc: 0.9609 - val_loss: 0.3058 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 943/3000\n",
      " - 0s - loss: 0.1540 - mean_squared_error: 0.0355 - acc: 0.9645 - val_loss: 0.3058 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 944/3000\n",
      " - 0s - loss: 0.1515 - mean_squared_error: 0.0337 - acc: 0.9670 - val_loss: 0.3057 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 945/3000\n",
      " - 0s - loss: 0.1497 - mean_squared_error: 0.0333 - acc: 0.9670 - val_loss: 0.3057 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 946/3000\n",
      " - 0s - loss: 0.1549 - mean_squared_error: 0.0351 - acc: 0.9621 - val_loss: 0.3057 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 947/3000\n",
      " - 0s - loss: 0.1515 - mean_squared_error: 0.0343 - acc: 0.9597 - val_loss: 0.3057 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 948/3000\n",
      " - 0s - loss: 0.1543 - mean_squared_error: 0.0352 - acc: 0.9609 - val_loss: 0.3056 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 949/3000\n",
      " - 0s - loss: 0.1505 - mean_squared_error: 0.0337 - acc: 0.9658 - val_loss: 0.3056 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 950/3000\n",
      " - 0s - loss: 0.1499 - mean_squared_error: 0.0333 - acc: 0.9670 - val_loss: 0.3056 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 951/3000\n",
      " - 0s - loss: 0.1571 - mean_squared_error: 0.0367 - acc: 0.9560 - val_loss: 0.3055 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 952/3000\n",
      " - 0s - loss: 0.1519 - mean_squared_error: 0.0341 - acc: 0.9621 - val_loss: 0.3055 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 953/3000\n",
      " - 0s - loss: 0.1516 - mean_squared_error: 0.0341 - acc: 0.9621 - val_loss: 0.3055 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 954/3000\n",
      " - 0s - loss: 0.1521 - mean_squared_error: 0.0342 - acc: 0.9670 - val_loss: 0.3054 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 955/3000\n",
      " - 0s - loss: 0.1551 - mean_squared_error: 0.0354 - acc: 0.9633 - val_loss: 0.3054 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 956/3000\n",
      " - 0s - loss: 0.1481 - mean_squared_error: 0.0328 - acc: 0.9621 - val_loss: 0.3054 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 957/3000\n",
      " - 0s - loss: 0.1515 - mean_squared_error: 0.0338 - acc: 0.9597 - val_loss: 0.3054 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 958/3000\n",
      " - 0s - loss: 0.1506 - mean_squared_error: 0.0338 - acc: 0.9645 - val_loss: 0.3053 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 959/3000\n",
      " - 0s - loss: 0.1513 - mean_squared_error: 0.0337 - acc: 0.9621 - val_loss: 0.3053 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 960/3000\n",
      " - 0s - loss: 0.1529 - mean_squared_error: 0.0352 - acc: 0.9560 - val_loss: 0.3053 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 961/3000\n",
      " - 0s - loss: 0.1531 - mean_squared_error: 0.0349 - acc: 0.9621 - val_loss: 0.3053 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 962/3000\n",
      " - 0s - loss: 0.1466 - mean_squared_error: 0.0321 - acc: 0.9645 - val_loss: 0.3053 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 963/3000\n",
      " - 0s - loss: 0.1522 - mean_squared_error: 0.0343 - acc: 0.9633 - val_loss: 0.3053 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 964/3000\n",
      " - 0s - loss: 0.1501 - mean_squared_error: 0.0339 - acc: 0.9621 - val_loss: 0.3053 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 965/3000\n",
      " - 0s - loss: 0.1477 - mean_squared_error: 0.0322 - acc: 0.9719 - val_loss: 0.3053 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 966/3000\n",
      " - 0s - loss: 0.1515 - mean_squared_error: 0.0339 - acc: 0.9682 - val_loss: 0.3052 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 967/3000\n",
      " - 0s - loss: 0.1526 - mean_squared_error: 0.0341 - acc: 0.9694 - val_loss: 0.3052 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 968/3000\n",
      " - 0s - loss: 0.1486 - mean_squared_error: 0.0332 - acc: 0.9682 - val_loss: 0.3052 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 969/3000\n",
      " - 0s - loss: 0.1487 - mean_squared_error: 0.0330 - acc: 0.9719 - val_loss: 0.3051 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 970/3000\n",
      " - 0s - loss: 0.1482 - mean_squared_error: 0.0328 - acc: 0.9658 - val_loss: 0.3051 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 971/3000\n",
      " - 0s - loss: 0.1526 - mean_squared_error: 0.0345 - acc: 0.9645 - val_loss: 0.3051 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 972/3000\n",
      " - 0s - loss: 0.1448 - mean_squared_error: 0.0318 - acc: 0.9658 - val_loss: 0.3050 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 973/3000\n",
      " - 0s - loss: 0.1479 - mean_squared_error: 0.0331 - acc: 0.9645 - val_loss: 0.3050 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 974/3000\n",
      " - 0s - loss: 0.1493 - mean_squared_error: 0.0335 - acc: 0.9670 - val_loss: 0.3050 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 975/3000\n",
      " - 0s - loss: 0.1505 - mean_squared_error: 0.0340 - acc: 0.9658 - val_loss: 0.3050 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 976/3000\n",
      " - 0s - loss: 0.1469 - mean_squared_error: 0.0329 - acc: 0.9670 - val_loss: 0.3049 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 977/3000\n",
      " - 0s - loss: 0.1454 - mean_squared_error: 0.0319 - acc: 0.9743 - val_loss: 0.3049 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 978/3000\n",
      " - 0s - loss: 0.1537 - mean_squared_error: 0.0355 - acc: 0.9584 - val_loss: 0.3049 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 979/3000\n",
      " - 0s - loss: 0.1522 - mean_squared_error: 0.0346 - acc: 0.9621 - val_loss: 0.3049 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 980/3000\n",
      " - 0s - loss: 0.1495 - mean_squared_error: 0.0338 - acc: 0.9658 - val_loss: 0.3049 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 981/3000\n",
      " - 0s - loss: 0.1509 - mean_squared_error: 0.0343 - acc: 0.9609 - val_loss: 0.3049 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 982/3000\n",
      " - 0s - loss: 0.1485 - mean_squared_error: 0.0329 - acc: 0.9731 - val_loss: 0.3049 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 983/3000\n",
      " - 0s - loss: 0.1467 - mean_squared_error: 0.0323 - acc: 0.9670 - val_loss: 0.3049 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 984/3000\n",
      " - 0s - loss: 0.1478 - mean_squared_error: 0.0332 - acc: 0.9621 - val_loss: 0.3049 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 985/3000\n",
      " - 0s - loss: 0.1524 - mean_squared_error: 0.0352 - acc: 0.9609 - val_loss: 0.3048 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 986/3000\n",
      " - 0s - loss: 0.1484 - mean_squared_error: 0.0329 - acc: 0.9670 - val_loss: 0.3048 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 987/3000\n",
      " - 0s - loss: 0.1431 - mean_squared_error: 0.0310 - acc: 0.9682 - val_loss: 0.3048 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 988/3000\n",
      " - 0s - loss: 0.1513 - mean_squared_error: 0.0339 - acc: 0.9707 - val_loss: 0.3048 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 989/3000\n",
      " - 0s - loss: 0.1473 - mean_squared_error: 0.0332 - acc: 0.9658 - val_loss: 0.3047 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 990/3000\n",
      " - 0s - loss: 0.1477 - mean_squared_error: 0.0327 - acc: 0.9633 - val_loss: 0.3047 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 991/3000\n",
      " - 0s - loss: 0.1457 - mean_squared_error: 0.0326 - acc: 0.9597 - val_loss: 0.3047 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 992/3000\n",
      " - 0s - loss: 0.1515 - mean_squared_error: 0.0342 - acc: 0.9609 - val_loss: 0.3047 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 993/3000\n",
      " - 0s - loss: 0.1465 - mean_squared_error: 0.0325 - acc: 0.9670 - val_loss: 0.3047 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 994/3000\n",
      " - 0s - loss: 0.1425 - mean_squared_error: 0.0311 - acc: 0.9694 - val_loss: 0.3047 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 995/3000\n",
      " - 0s - loss: 0.1494 - mean_squared_error: 0.0337 - acc: 0.9633 - val_loss: 0.3046 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 996/3000\n",
      " - 0s - loss: 0.1460 - mean_squared_error: 0.0322 - acc: 0.9658 - val_loss: 0.3046 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 997/3000\n",
      " - 0s - loss: 0.1419 - mean_squared_error: 0.0302 - acc: 0.9707 - val_loss: 0.3046 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 998/3000\n",
      " - 0s - loss: 0.1458 - mean_squared_error: 0.0322 - acc: 0.9670 - val_loss: 0.3046 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 999/3000\n",
      " - 0s - loss: 0.1423 - mean_squared_error: 0.0311 - acc: 0.9621 - val_loss: 0.3046 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 1000/3000\n",
      " - 0s - loss: 0.1405 - mean_squared_error: 0.0306 - acc: 0.9682 - val_loss: 0.3046 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 1001/3000\n",
      " - 0s - loss: 0.1405 - mean_squared_error: 0.0303 - acc: 0.9707 - val_loss: 0.3046 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 1002/3000\n",
      " - 0s - loss: 0.1432 - mean_squared_error: 0.0316 - acc: 0.9645 - val_loss: 0.3046 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 1003/3000\n",
      " - 0s - loss: 0.1441 - mean_squared_error: 0.0320 - acc: 0.9621 - val_loss: 0.3046 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 1004/3000\n",
      " - 0s - loss: 0.1437 - mean_squared_error: 0.0317 - acc: 0.9694 - val_loss: 0.3046 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 1005/3000\n",
      " - 0s - loss: 0.1450 - mean_squared_error: 0.0327 - acc: 0.9633 - val_loss: 0.3046 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 1006/3000\n",
      " - 0s - loss: 0.1478 - mean_squared_error: 0.0328 - acc: 0.9694 - val_loss: 0.3046 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 1007/3000\n",
      " - 0s - loss: 0.1482 - mean_squared_error: 0.0338 - acc: 0.9609 - val_loss: 0.3045 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 1008/3000\n",
      " - 0s - loss: 0.1428 - mean_squared_error: 0.0313 - acc: 0.9694 - val_loss: 0.3045 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 1009/3000\n",
      " - 0s - loss: 0.1449 - mean_squared_error: 0.0319 - acc: 0.9694 - val_loss: 0.3045 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 1010/3000\n",
      " - 0s - loss: 0.1453 - mean_squared_error: 0.0328 - acc: 0.9597 - val_loss: 0.3045 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 1011/3000\n",
      " - 0s - loss: 0.1437 - mean_squared_error: 0.0317 - acc: 0.9645 - val_loss: 0.3044 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 1012/3000\n",
      " - 0s - loss: 0.1462 - mean_squared_error: 0.0322 - acc: 0.9682 - val_loss: 0.3044 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 1013/3000\n",
      " - 0s - loss: 0.1439 - mean_squared_error: 0.0321 - acc: 0.9682 - val_loss: 0.3044 - val_mean_squared_error: 0.0930 - val_acc: 0.8613\n",
      "Epoch 1014/3000\n",
      " - 0s - loss: 0.1448 - mean_squared_error: 0.0324 - acc: 0.9633 - val_loss: 0.3044 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1015/3000\n",
      " - 0s - loss: 0.1469 - mean_squared_error: 0.0334 - acc: 0.9597 - val_loss: 0.3044 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1016/3000\n",
      " - 0s - loss: 0.1503 - mean_squared_error: 0.0345 - acc: 0.9621 - val_loss: 0.3044 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1017/3000\n",
      " - 0s - loss: 0.1449 - mean_squared_error: 0.0323 - acc: 0.9682 - val_loss: 0.3044 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1018/3000\n",
      " - 0s - loss: 0.1441 - mean_squared_error: 0.0323 - acc: 0.9682 - val_loss: 0.3043 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1019/3000\n",
      " - 0s - loss: 0.1465 - mean_squared_error: 0.0333 - acc: 0.9633 - val_loss: 0.3043 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1020/3000\n",
      " - 0s - loss: 0.1407 - mean_squared_error: 0.0310 - acc: 0.9645 - val_loss: 0.3043 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1021/3000\n",
      " - 0s - loss: 0.1399 - mean_squared_error: 0.0307 - acc: 0.9682 - val_loss: 0.3043 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1022/3000\n",
      " - 0s - loss: 0.1384 - mean_squared_error: 0.0299 - acc: 0.9682 - val_loss: 0.3043 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1023/3000\n",
      " - 0s - loss: 0.1450 - mean_squared_error: 0.0323 - acc: 0.9682 - val_loss: 0.3043 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1024/3000\n",
      " - 0s - loss: 0.1429 - mean_squared_error: 0.0321 - acc: 0.9658 - val_loss: 0.3043 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1025/3000\n",
      " - 0s - loss: 0.1399 - mean_squared_error: 0.0304 - acc: 0.9743 - val_loss: 0.3043 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1026/3000\n",
      " - 0s - loss: 0.1472 - mean_squared_error: 0.0332 - acc: 0.9621 - val_loss: 0.3043 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1027/3000\n",
      " - 0s - loss: 0.1397 - mean_squared_error: 0.0306 - acc: 0.9670 - val_loss: 0.3043 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1028/3000\n",
      " - 0s - loss: 0.1438 - mean_squared_error: 0.0322 - acc: 0.9682 - val_loss: 0.3043 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1029/3000\n",
      " - 0s - loss: 0.1385 - mean_squared_error: 0.0302 - acc: 0.9682 - val_loss: 0.3043 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1030/3000\n",
      " - 0s - loss: 0.1414 - mean_squared_error: 0.0307 - acc: 0.9707 - val_loss: 0.3042 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1031/3000\n",
      " - 0s - loss: 0.1377 - mean_squared_error: 0.0300 - acc: 0.9731 - val_loss: 0.3042 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1032/3000\n",
      " - 0s - loss: 0.1428 - mean_squared_error: 0.0319 - acc: 0.9682 - val_loss: 0.3042 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1033/3000\n",
      " - 0s - loss: 0.1415 - mean_squared_error: 0.0314 - acc: 0.9670 - val_loss: 0.3042 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1034/3000\n",
      " - 0s - loss: 0.1421 - mean_squared_error: 0.0318 - acc: 0.9621 - val_loss: 0.3042 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1035/3000\n",
      " - 0s - loss: 0.1410 - mean_squared_error: 0.0311 - acc: 0.9670 - val_loss: 0.3041 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1036/3000\n",
      " - 0s - loss: 0.1374 - mean_squared_error: 0.0296 - acc: 0.9694 - val_loss: 0.3041 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1037/3000\n",
      " - 0s - loss: 0.1402 - mean_squared_error: 0.0311 - acc: 0.9694 - val_loss: 0.3041 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1038/3000\n",
      " - 0s - loss: 0.1390 - mean_squared_error: 0.0309 - acc: 0.9633 - val_loss: 0.3041 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1039/3000\n",
      " - 0s - loss: 0.1415 - mean_squared_error: 0.0319 - acc: 0.9670 - val_loss: 0.3041 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1040/3000\n",
      " - 0s - loss: 0.1406 - mean_squared_error: 0.0312 - acc: 0.9682 - val_loss: 0.3041 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1041/3000\n",
      " - 0s - loss: 0.1386 - mean_squared_error: 0.0307 - acc: 0.9682 - val_loss: 0.3041 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1042/3000\n",
      " - 0s - loss: 0.1379 - mean_squared_error: 0.0303 - acc: 0.9670 - val_loss: 0.3041 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1043/3000\n",
      " - 0s - loss: 0.1355 - mean_squared_error: 0.0296 - acc: 0.9670 - val_loss: 0.3041 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1044/3000\n",
      " - 0s - loss: 0.1419 - mean_squared_error: 0.0318 - acc: 0.9682 - val_loss: 0.3041 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1045/3000\n",
      " - 0s - loss: 0.1409 - mean_squared_error: 0.0317 - acc: 0.9633 - val_loss: 0.3041 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1046/3000\n",
      " - 0s - loss: 0.1406 - mean_squared_error: 0.0307 - acc: 0.9719 - val_loss: 0.3041 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1047/3000\n",
      " - 0s - loss: 0.1398 - mean_squared_error: 0.0304 - acc: 0.9719 - val_loss: 0.3041 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1048/3000\n",
      " - 0s - loss: 0.1392 - mean_squared_error: 0.0307 - acc: 0.9694 - val_loss: 0.3040 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1049/3000\n",
      " - 0s - loss: 0.1395 - mean_squared_error: 0.0307 - acc: 0.9670 - val_loss: 0.3040 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1050/3000\n",
      " - 0s - loss: 0.1435 - mean_squared_error: 0.0323 - acc: 0.9633 - val_loss: 0.3040 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1051/3000\n",
      " - 0s - loss: 0.1396 - mean_squared_error: 0.0307 - acc: 0.9682 - val_loss: 0.3040 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1052/3000\n",
      " - 0s - loss: 0.1367 - mean_squared_error: 0.0301 - acc: 0.9658 - val_loss: 0.3040 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1053/3000\n",
      " - 0s - loss: 0.1351 - mean_squared_error: 0.0290 - acc: 0.9682 - val_loss: 0.3039 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1054/3000\n",
      " - 0s - loss: 0.1408 - mean_squared_error: 0.0315 - acc: 0.9645 - val_loss: 0.3039 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1055/3000\n",
      " - 0s - loss: 0.1416 - mean_squared_error: 0.0315 - acc: 0.9645 - val_loss: 0.3039 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1056/3000\n",
      " - 0s - loss: 0.1419 - mean_squared_error: 0.0317 - acc: 0.9670 - val_loss: 0.3039 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1057/3000\n",
      " - 0s - loss: 0.1423 - mean_squared_error: 0.0321 - acc: 0.9707 - val_loss: 0.3039 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1058/3000\n",
      " - 0s - loss: 0.1379 - mean_squared_error: 0.0302 - acc: 0.9682 - val_loss: 0.3039 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1059/3000\n",
      " - 0s - loss: 0.1385 - mean_squared_error: 0.0307 - acc: 0.9658 - val_loss: 0.3039 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1060/3000\n",
      " - 0s - loss: 0.1372 - mean_squared_error: 0.0304 - acc: 0.9658 - val_loss: 0.3039 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1061/3000\n",
      " - 0s - loss: 0.1373 - mean_squared_error: 0.0300 - acc: 0.9694 - val_loss: 0.3039 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1062/3000\n",
      " - 0s - loss: 0.1403 - mean_squared_error: 0.0311 - acc: 0.9694 - val_loss: 0.3039 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1063/3000\n",
      " - 0s - loss: 0.1410 - mean_squared_error: 0.0315 - acc: 0.9645 - val_loss: 0.3039 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1064/3000\n",
      " - 0s - loss: 0.1373 - mean_squared_error: 0.0305 - acc: 0.9670 - val_loss: 0.3039 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1065/3000\n",
      " - 0s - loss: 0.1411 - mean_squared_error: 0.0314 - acc: 0.9707 - val_loss: 0.3038 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1066/3000\n",
      " - 0s - loss: 0.1396 - mean_squared_error: 0.0304 - acc: 0.9707 - val_loss: 0.3038 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1067/3000\n",
      " - 0s - loss: 0.1365 - mean_squared_error: 0.0300 - acc: 0.9719 - val_loss: 0.3038 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1068/3000\n",
      " - 0s - loss: 0.1431 - mean_squared_error: 0.0326 - acc: 0.9645 - val_loss: 0.3038 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1069/3000\n",
      " - 0s - loss: 0.1344 - mean_squared_error: 0.0291 - acc: 0.9694 - val_loss: 0.3038 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1070/3000\n",
      " - 0s - loss: 0.1393 - mean_squared_error: 0.0314 - acc: 0.9645 - val_loss: 0.3038 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1071/3000\n",
      " - 0s - loss: 0.1361 - mean_squared_error: 0.0297 - acc: 0.9658 - val_loss: 0.3038 - val_mean_squared_error: 0.0931 - val_acc: 0.8613\n",
      "Epoch 1072/3000\n",
      " - 0s - loss: 0.1362 - mean_squared_error: 0.0301 - acc: 0.9707 - val_loss: 0.3037 - val_mean_squared_error: 0.0931 - val_acc: 0.8650\n",
      "Epoch 1073/3000\n",
      " - 0s - loss: 0.1411 - mean_squared_error: 0.0318 - acc: 0.9670 - val_loss: 0.3037 - val_mean_squared_error: 0.0931 - val_acc: 0.8650\n",
      "Epoch 1074/3000\n",
      " - 0s - loss: 0.1373 - mean_squared_error: 0.0303 - acc: 0.9682 - val_loss: 0.3037 - val_mean_squared_error: 0.0931 - val_acc: 0.8650\n",
      "Epoch 1075/3000\n",
      " - 0s - loss: 0.1388 - mean_squared_error: 0.0310 - acc: 0.9707 - val_loss: 0.3037 - val_mean_squared_error: 0.0931 - val_acc: 0.8650\n",
      "Epoch 1076/3000\n",
      " - 0s - loss: 0.1396 - mean_squared_error: 0.0313 - acc: 0.9633 - val_loss: 0.3037 - val_mean_squared_error: 0.0931 - val_acc: 0.8650\n",
      "Epoch 1077/3000\n",
      " - 0s - loss: 0.1378 - mean_squared_error: 0.0308 - acc: 0.9658 - val_loss: 0.3037 - val_mean_squared_error: 0.0931 - val_acc: 0.8650\n",
      "Epoch 1078/3000\n",
      " - 0s - loss: 0.1304 - mean_squared_error: 0.0276 - acc: 0.9768 - val_loss: 0.3037 - val_mean_squared_error: 0.0931 - val_acc: 0.8650\n",
      "Epoch 1079/3000\n",
      " - 0s - loss: 0.1347 - mean_squared_error: 0.0294 - acc: 0.9670 - val_loss: 0.3037 - val_mean_squared_error: 0.0931 - val_acc: 0.8650\n",
      "Epoch 1080/3000\n",
      " - 0s - loss: 0.1324 - mean_squared_error: 0.0283 - acc: 0.9719 - val_loss: 0.3037 - val_mean_squared_error: 0.0931 - val_acc: 0.8650\n",
      "Epoch 1081/3000\n",
      " - 0s - loss: 0.1387 - mean_squared_error: 0.0310 - acc: 0.9633 - val_loss: 0.3037 - val_mean_squared_error: 0.0931 - val_acc: 0.8650\n",
      "Epoch 1082/3000\n",
      " - 0s - loss: 0.1341 - mean_squared_error: 0.0291 - acc: 0.9682 - val_loss: 0.3037 - val_mean_squared_error: 0.0931 - val_acc: 0.8650\n",
      "Epoch 1083/3000\n",
      " - 0s - loss: 0.1365 - mean_squared_error: 0.0304 - acc: 0.9694 - val_loss: 0.3037 - val_mean_squared_error: 0.0931 - val_acc: 0.8650\n",
      "Epoch 1084/3000\n",
      " - 0s - loss: 0.1341 - mean_squared_error: 0.0294 - acc: 0.9707 - val_loss: 0.3036 - val_mean_squared_error: 0.0931 - val_acc: 0.8650\n",
      "Epoch 1085/3000\n",
      " - 0s - loss: 0.1332 - mean_squared_error: 0.0289 - acc: 0.9731 - val_loss: 0.3036 - val_mean_squared_error: 0.0931 - val_acc: 0.8650\n",
      "Epoch 1086/3000\n",
      " - 0s - loss: 0.1327 - mean_squared_error: 0.0289 - acc: 0.9682 - val_loss: 0.3036 - val_mean_squared_error: 0.0931 - val_acc: 0.8650\n",
      "Epoch 1087/3000\n",
      " - 0s - loss: 0.1309 - mean_squared_error: 0.0282 - acc: 0.9682 - val_loss: 0.3036 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1088/3000\n",
      " - 0s - loss: 0.1358 - mean_squared_error: 0.0303 - acc: 0.9645 - val_loss: 0.3036 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1089/3000\n",
      " - 0s - loss: 0.1382 - mean_squared_error: 0.0309 - acc: 0.9707 - val_loss: 0.3036 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1090/3000\n",
      " - 0s - loss: 0.1372 - mean_squared_error: 0.0301 - acc: 0.9694 - val_loss: 0.3036 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1091/3000\n",
      " - 0s - loss: 0.1361 - mean_squared_error: 0.0303 - acc: 0.9670 - val_loss: 0.3036 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1092/3000\n",
      " - 0s - loss: 0.1334 - mean_squared_error: 0.0289 - acc: 0.9719 - val_loss: 0.3036 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1093/3000\n",
      " - 0s - loss: 0.1351 - mean_squared_error: 0.0298 - acc: 0.9707 - val_loss: 0.3035 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1094/3000\n",
      " - 0s - loss: 0.1384 - mean_squared_error: 0.0310 - acc: 0.9633 - val_loss: 0.3035 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1095/3000\n",
      " - 0s - loss: 0.1387 - mean_squared_error: 0.0310 - acc: 0.9682 - val_loss: 0.3035 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1096/3000\n",
      " - 0s - loss: 0.1299 - mean_squared_error: 0.0281 - acc: 0.9719 - val_loss: 0.3035 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1097/3000\n",
      " - 0s - loss: 0.1354 - mean_squared_error: 0.0295 - acc: 0.9707 - val_loss: 0.3035 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1098/3000\n",
      " - 0s - loss: 0.1329 - mean_squared_error: 0.0292 - acc: 0.9645 - val_loss: 0.3035 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1099/3000\n",
      " - 0s - loss: 0.1351 - mean_squared_error: 0.0295 - acc: 0.9707 - val_loss: 0.3035 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1100/3000\n",
      " - 0s - loss: 0.1348 - mean_squared_error: 0.0290 - acc: 0.9780 - val_loss: 0.3035 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1101/3000\n",
      " - 0s - loss: 0.1321 - mean_squared_error: 0.0288 - acc: 0.9670 - val_loss: 0.3035 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1102/3000\n",
      " - 0s - loss: 0.1282 - mean_squared_error: 0.0272 - acc: 0.9756 - val_loss: 0.3035 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1103/3000\n",
      " - 0s - loss: 0.1335 - mean_squared_error: 0.0287 - acc: 0.9743 - val_loss: 0.3035 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1104/3000\n",
      " - 0s - loss: 0.1359 - mean_squared_error: 0.0303 - acc: 0.9633 - val_loss: 0.3035 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1105/3000\n",
      " - 0s - loss: 0.1270 - mean_squared_error: 0.0271 - acc: 0.9682 - val_loss: 0.3035 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1106/3000\n",
      " - 0s - loss: 0.1331 - mean_squared_error: 0.0290 - acc: 0.9694 - val_loss: 0.3035 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1107/3000\n",
      " - 0s - loss: 0.1337 - mean_squared_error: 0.0293 - acc: 0.9719 - val_loss: 0.3035 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1108/3000\n",
      " - 0s - loss: 0.1383 - mean_squared_error: 0.0304 - acc: 0.9731 - val_loss: 0.3035 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1109/3000\n",
      " - 0s - loss: 0.1324 - mean_squared_error: 0.0290 - acc: 0.9694 - val_loss: 0.3035 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1110/3000\n",
      " - 0s - loss: 0.1297 - mean_squared_error: 0.0283 - acc: 0.9731 - val_loss: 0.3035 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1111/3000\n",
      " - 0s - loss: 0.1286 - mean_squared_error: 0.0269 - acc: 0.9780 - val_loss: 0.3035 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1112/3000\n",
      " - 0s - loss: 0.1363 - mean_squared_error: 0.0300 - acc: 0.9645 - val_loss: 0.3034 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1113/3000\n",
      " - 0s - loss: 0.1334 - mean_squared_error: 0.0291 - acc: 0.9645 - val_loss: 0.3034 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1114/3000\n",
      " - 0s - loss: 0.1330 - mean_squared_error: 0.0292 - acc: 0.9707 - val_loss: 0.3034 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1115/3000\n",
      " - 0s - loss: 0.1362 - mean_squared_error: 0.0307 - acc: 0.9621 - val_loss: 0.3034 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1116/3000\n",
      " - 0s - loss: 0.1326 - mean_squared_error: 0.0288 - acc: 0.9743 - val_loss: 0.3034 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1117/3000\n",
      " - 0s - loss: 0.1340 - mean_squared_error: 0.0299 - acc: 0.9682 - val_loss: 0.3034 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1118/3000\n",
      " - 0s - loss: 0.1323 - mean_squared_error: 0.0288 - acc: 0.9621 - val_loss: 0.3034 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1119/3000\n",
      " - 0s - loss: 0.1364 - mean_squared_error: 0.0302 - acc: 0.9682 - val_loss: 0.3034 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1120/3000\n",
      " - 0s - loss: 0.1271 - mean_squared_error: 0.0270 - acc: 0.9743 - val_loss: 0.3034 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1121/3000\n",
      " - 0s - loss: 0.1283 - mean_squared_error: 0.0278 - acc: 0.9743 - val_loss: 0.3034 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1122/3000\n",
      " - 0s - loss: 0.1349 - mean_squared_error: 0.0299 - acc: 0.9633 - val_loss: 0.3034 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1123/3000\n",
      " - 0s - loss: 0.1333 - mean_squared_error: 0.0299 - acc: 0.9645 - val_loss: 0.3034 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1124/3000\n",
      " - 0s - loss: 0.1263 - mean_squared_error: 0.0265 - acc: 0.9731 - val_loss: 0.3034 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1125/3000\n",
      " - 0s - loss: 0.1297 - mean_squared_error: 0.0281 - acc: 0.9670 - val_loss: 0.3034 - val_mean_squared_error: 0.0932 - val_acc: 0.8650\n",
      "Epoch 1126/3000\n",
      " - 0s - loss: 0.1335 - mean_squared_error: 0.0294 - acc: 0.9658 - val_loss: 0.3034 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1127/3000\n",
      " - 0s - loss: 0.1303 - mean_squared_error: 0.0284 - acc: 0.9719 - val_loss: 0.3034 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1128/3000\n",
      " - 0s - loss: 0.1261 - mean_squared_error: 0.0268 - acc: 0.9719 - val_loss: 0.3034 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1129/3000\n",
      " - 0s - loss: 0.1335 - mean_squared_error: 0.0296 - acc: 0.9670 - val_loss: 0.3034 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1130/3000\n",
      " - 0s - loss: 0.1285 - mean_squared_error: 0.0279 - acc: 0.9694 - val_loss: 0.3034 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1131/3000\n",
      " - 0s - loss: 0.1339 - mean_squared_error: 0.0299 - acc: 0.9670 - val_loss: 0.3034 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1132/3000\n",
      " - 0s - loss: 0.1329 - mean_squared_error: 0.0295 - acc: 0.9707 - val_loss: 0.3034 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1133/3000\n",
      " - 0s - loss: 0.1249 - mean_squared_error: 0.0265 - acc: 0.9719 - val_loss: 0.3034 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1134/3000\n",
      " - 0s - loss: 0.1275 - mean_squared_error: 0.0270 - acc: 0.9731 - val_loss: 0.3034 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1135/3000\n",
      " - 0s - loss: 0.1248 - mean_squared_error: 0.0266 - acc: 0.9780 - val_loss: 0.3033 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1136/3000\n",
      " - 0s - loss: 0.1305 - mean_squared_error: 0.0278 - acc: 0.9743 - val_loss: 0.3033 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1137/3000\n",
      " - 0s - loss: 0.1283 - mean_squared_error: 0.0273 - acc: 0.9731 - val_loss: 0.3033 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1138/3000\n",
      " - 0s - loss: 0.1323 - mean_squared_error: 0.0294 - acc: 0.9719 - val_loss: 0.3033 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1139/3000\n",
      " - 0s - loss: 0.1318 - mean_squared_error: 0.0289 - acc: 0.9682 - val_loss: 0.3033 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1140/3000\n",
      " - 0s - loss: 0.1318 - mean_squared_error: 0.0291 - acc: 0.9707 - val_loss: 0.3033 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1141/3000\n",
      " - 0s - loss: 0.1250 - mean_squared_error: 0.0266 - acc: 0.9719 - val_loss: 0.3033 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1142/3000\n",
      " - 0s - loss: 0.1301 - mean_squared_error: 0.0282 - acc: 0.9731 - val_loss: 0.3033 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1143/3000\n",
      " - 0s - loss: 0.1326 - mean_squared_error: 0.0293 - acc: 0.9682 - val_loss: 0.3033 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1144/3000\n",
      " - 0s - loss: 0.1232 - mean_squared_error: 0.0255 - acc: 0.9768 - val_loss: 0.3033 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1145/3000\n",
      " - 0s - loss: 0.1275 - mean_squared_error: 0.0279 - acc: 0.9682 - val_loss: 0.3033 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1146/3000\n",
      " - 0s - loss: 0.1284 - mean_squared_error: 0.0281 - acc: 0.9707 - val_loss: 0.3033 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1147/3000\n",
      " - 0s - loss: 0.1360 - mean_squared_error: 0.0306 - acc: 0.9645 - val_loss: 0.3033 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1148/3000\n",
      " - 0s - loss: 0.1249 - mean_squared_error: 0.0264 - acc: 0.9743 - val_loss: 0.3033 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1149/3000\n",
      " - 0s - loss: 0.1297 - mean_squared_error: 0.0282 - acc: 0.9707 - val_loss: 0.3033 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1150/3000\n",
      " - 0s - loss: 0.1281 - mean_squared_error: 0.0277 - acc: 0.9694 - val_loss: 0.3033 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1151/3000\n",
      " - 0s - loss: 0.1341 - mean_squared_error: 0.0304 - acc: 0.9609 - val_loss: 0.3033 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1152/3000\n",
      " - 0s - loss: 0.1294 - mean_squared_error: 0.0283 - acc: 0.9719 - val_loss: 0.3033 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1153/3000\n",
      " - 0s - loss: 0.1287 - mean_squared_error: 0.0277 - acc: 0.9743 - val_loss: 0.3033 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1154/3000\n",
      " - 0s - loss: 0.1236 - mean_squared_error: 0.0260 - acc: 0.9707 - val_loss: 0.3033 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1155/3000\n",
      " - 0s - loss: 0.1280 - mean_squared_error: 0.0280 - acc: 0.9682 - val_loss: 0.3033 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1156/3000\n",
      " - 0s - loss: 0.1323 - mean_squared_error: 0.0299 - acc: 0.9694 - val_loss: 0.3033 - val_mean_squared_error: 0.0933 - val_acc: 0.8650\n",
      "Epoch 1157/3000\n",
      " - 0s - loss: 0.1262 - mean_squared_error: 0.0277 - acc: 0.9743 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1158/3000\n",
      " - 0s - loss: 0.1299 - mean_squared_error: 0.0288 - acc: 0.9658 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1159/3000\n",
      " - 0s - loss: 0.1300 - mean_squared_error: 0.0287 - acc: 0.9731 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1160/3000\n",
      " - 0s - loss: 0.1300 - mean_squared_error: 0.0286 - acc: 0.9743 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1161/3000\n",
      " - 0s - loss: 0.1274 - mean_squared_error: 0.0278 - acc: 0.9756 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1162/3000\n",
      " - 0s - loss: 0.1296 - mean_squared_error: 0.0282 - acc: 0.9694 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1163/3000\n",
      " - 0s - loss: 0.1291 - mean_squared_error: 0.0281 - acc: 0.9694 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1164/3000\n",
      " - 0s - loss: 0.1276 - mean_squared_error: 0.0272 - acc: 0.9731 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1165/3000\n",
      " - 0s - loss: 0.1280 - mean_squared_error: 0.0276 - acc: 0.9694 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1166/3000\n",
      " - 0s - loss: 0.1231 - mean_squared_error: 0.0262 - acc: 0.9731 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1167/3000\n",
      " - 0s - loss: 0.1254 - mean_squared_error: 0.0265 - acc: 0.9780 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1168/3000\n",
      " - 0s - loss: 0.1252 - mean_squared_error: 0.0272 - acc: 0.9731 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1169/3000\n",
      " - 0s - loss: 0.1237 - mean_squared_error: 0.0262 - acc: 0.9743 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1170/3000\n",
      " - 0s - loss: 0.1335 - mean_squared_error: 0.0300 - acc: 0.9682 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1171/3000\n",
      " - 0s - loss: 0.1275 - mean_squared_error: 0.0278 - acc: 0.9719 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1172/3000\n",
      " - 0s - loss: 0.1268 - mean_squared_error: 0.0278 - acc: 0.9658 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1173/3000\n",
      " - 0s - loss: 0.1285 - mean_squared_error: 0.0279 - acc: 0.9719 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1174/3000\n",
      " - 0s - loss: 0.1213 - mean_squared_error: 0.0259 - acc: 0.9694 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1175/3000\n",
      " - 0s - loss: 0.1226 - mean_squared_error: 0.0262 - acc: 0.9682 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1176/3000\n",
      " - 0s - loss: 0.1176 - mean_squared_error: 0.0243 - acc: 0.9743 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1177/3000\n",
      " - 0s - loss: 0.1268 - mean_squared_error: 0.0280 - acc: 0.9719 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1178/3000\n",
      " - 0s - loss: 0.1229 - mean_squared_error: 0.0262 - acc: 0.9731 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1179/3000\n",
      " - 0s - loss: 0.1251 - mean_squared_error: 0.0282 - acc: 0.9670 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1180/3000\n",
      " - 0s - loss: 0.1274 - mean_squared_error: 0.0275 - acc: 0.9731 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1181/3000\n",
      " - 0s - loss: 0.1267 - mean_squared_error: 0.0277 - acc: 0.9694 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1182/3000\n",
      " - 0s - loss: 0.1235 - mean_squared_error: 0.0264 - acc: 0.9719 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1183/3000\n",
      " - 0s - loss: 0.1251 - mean_squared_error: 0.0271 - acc: 0.9719 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1184/3000\n",
      " - 0s - loss: 0.1273 - mean_squared_error: 0.0279 - acc: 0.9670 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1185/3000\n",
      " - 0s - loss: 0.1289 - mean_squared_error: 0.0283 - acc: 0.9694 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1186/3000\n",
      " - 0s - loss: 0.1287 - mean_squared_error: 0.0282 - acc: 0.9756 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1187/3000\n",
      " - 0s - loss: 0.1216 - mean_squared_error: 0.0257 - acc: 0.9719 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1188/3000\n",
      " - 0s - loss: 0.1208 - mean_squared_error: 0.0252 - acc: 0.9792 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1189/3000\n",
      " - 0s - loss: 0.1226 - mean_squared_error: 0.0260 - acc: 0.9731 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1190/3000\n",
      " - 0s - loss: 0.1251 - mean_squared_error: 0.0268 - acc: 0.9731 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1191/3000\n",
      " - 0s - loss: 0.1257 - mean_squared_error: 0.0272 - acc: 0.9731 - val_loss: 0.3033 - val_mean_squared_error: 0.0934 - val_acc: 0.8650\n",
      "Epoch 1192/3000\n",
      " - 0s - loss: 0.1221 - mean_squared_error: 0.0253 - acc: 0.9780 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n",
      "Epoch 1193/3000\n",
      " - 0s - loss: 0.1292 - mean_squared_error: 0.0282 - acc: 0.9694 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n",
      "Epoch 1194/3000\n",
      " - 0s - loss: 0.1253 - mean_squared_error: 0.0271 - acc: 0.9743 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n",
      "Epoch 1195/3000\n",
      " - 0s - loss: 0.1222 - mean_squared_error: 0.0260 - acc: 0.9743 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n",
      "Epoch 1196/3000\n",
      " - 0s - loss: 0.1219 - mean_squared_error: 0.0260 - acc: 0.9780 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n",
      "Epoch 1197/3000\n",
      " - 0s - loss: 0.1253 - mean_squared_error: 0.0267 - acc: 0.9719 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n",
      "Epoch 1198/3000\n",
      " - 0s - loss: 0.1232 - mean_squared_error: 0.0263 - acc: 0.9743 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n",
      "Epoch 1199/3000\n",
      " - 0s - loss: 0.1224 - mean_squared_error: 0.0260 - acc: 0.9719 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n",
      "Epoch 1200/3000\n",
      " - 0s - loss: 0.1202 - mean_squared_error: 0.0253 - acc: 0.9743 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n",
      "Epoch 1201/3000\n",
      " - 0s - loss: 0.1178 - mean_squared_error: 0.0250 - acc: 0.9731 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n",
      "Epoch 1202/3000\n",
      " - 0s - loss: 0.1243 - mean_squared_error: 0.0270 - acc: 0.9682 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n",
      "Epoch 1203/3000\n",
      " - 0s - loss: 0.1149 - mean_squared_error: 0.0239 - acc: 0.9756 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n",
      "Epoch 1204/3000\n",
      " - 0s - loss: 0.1253 - mean_squared_error: 0.0268 - acc: 0.9707 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n",
      "Epoch 1205/3000\n",
      " - 0s - loss: 0.1258 - mean_squared_error: 0.0270 - acc: 0.9768 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n",
      "Epoch 1206/3000\n",
      " - 0s - loss: 0.1198 - mean_squared_error: 0.0250 - acc: 0.9756 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n",
      "Epoch 1207/3000\n",
      " - 0s - loss: 0.1167 - mean_squared_error: 0.0243 - acc: 0.9792 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n",
      "Epoch 1208/3000\n",
      " - 0s - loss: 0.1184 - mean_squared_error: 0.0246 - acc: 0.9792 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n",
      "Epoch 1209/3000\n",
      " - 0s - loss: 0.1173 - mean_squared_error: 0.0243 - acc: 0.9768 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n",
      "Epoch 1210/3000\n",
      " - 0s - loss: 0.1242 - mean_squared_error: 0.0271 - acc: 0.9719 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n",
      "Epoch 1211/3000\n",
      " - 0s - loss: 0.1249 - mean_squared_error: 0.0267 - acc: 0.9780 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n",
      "Epoch 1212/3000\n",
      " - 0s - loss: 0.1250 - mean_squared_error: 0.0276 - acc: 0.9719 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1213/3000\n",
      " - 0s - loss: 0.1246 - mean_squared_error: 0.0274 - acc: 0.9694 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n",
      "Epoch 1214/3000\n",
      " - 0s - loss: 0.1234 - mean_squared_error: 0.0268 - acc: 0.9731 - val_loss: 0.3033 - val_mean_squared_error: 0.0935 - val_acc: 0.8650\n",
      "Epoch 1215/3000\n",
      " - 0s - loss: 0.1227 - mean_squared_error: 0.0268 - acc: 0.9694 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1216/3000\n",
      " - 0s - loss: 0.1221 - mean_squared_error: 0.0265 - acc: 0.9731 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1217/3000\n",
      " - 0s - loss: 0.1212 - mean_squared_error: 0.0257 - acc: 0.9768 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1218/3000\n",
      " - 0s - loss: 0.1239 - mean_squared_error: 0.0265 - acc: 0.9707 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1219/3000\n",
      " - 0s - loss: 0.1196 - mean_squared_error: 0.0256 - acc: 0.9731 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1220/3000\n",
      " - 0s - loss: 0.1175 - mean_squared_error: 0.0246 - acc: 0.9743 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1221/3000\n",
      " - 0s - loss: 0.1204 - mean_squared_error: 0.0255 - acc: 0.9682 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1222/3000\n",
      " - 0s - loss: 0.1189 - mean_squared_error: 0.0250 - acc: 0.9780 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1223/3000\n",
      " - 0s - loss: 0.1214 - mean_squared_error: 0.0263 - acc: 0.9743 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1224/3000\n",
      " - 0s - loss: 0.1214 - mean_squared_error: 0.0260 - acc: 0.9792 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1225/3000\n",
      " - 0s - loss: 0.1214 - mean_squared_error: 0.0264 - acc: 0.9743 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1226/3000\n",
      " - 0s - loss: 0.1213 - mean_squared_error: 0.0259 - acc: 0.9756 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1227/3000\n",
      " - 0s - loss: 0.1213 - mean_squared_error: 0.0266 - acc: 0.9731 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1228/3000\n",
      " - 0s - loss: 0.1197 - mean_squared_error: 0.0260 - acc: 0.9694 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1229/3000\n",
      " - 0s - loss: 0.1243 - mean_squared_error: 0.0273 - acc: 0.9658 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1230/3000\n",
      " - 0s - loss: 0.1114 - mean_squared_error: 0.0221 - acc: 0.9804 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1231/3000\n",
      " - 0s - loss: 0.1247 - mean_squared_error: 0.0274 - acc: 0.9707 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1232/3000\n",
      " - 0s - loss: 0.1168 - mean_squared_error: 0.0245 - acc: 0.9707 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1233/3000\n",
      " - 0s - loss: 0.1219 - mean_squared_error: 0.0262 - acc: 0.9743 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1234/3000\n",
      " - 0s - loss: 0.1214 - mean_squared_error: 0.0267 - acc: 0.9694 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1235/3000\n",
      " - 0s - loss: 0.1200 - mean_squared_error: 0.0256 - acc: 0.9743 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1236/3000\n",
      " - 0s - loss: 0.1199 - mean_squared_error: 0.0254 - acc: 0.9743 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1237/3000\n",
      " - 0s - loss: 0.1144 - mean_squared_error: 0.0238 - acc: 0.9756 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1238/3000\n",
      " - 0s - loss: 0.1205 - mean_squared_error: 0.0256 - acc: 0.9719 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1239/3000\n",
      " - 0s - loss: 0.1204 - mean_squared_error: 0.0256 - acc: 0.9756 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1240/3000\n",
      " - 0s - loss: 0.1189 - mean_squared_error: 0.0257 - acc: 0.9719 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "Epoch 1241/3000\n",
      " - 0s - loss: 0.1174 - mean_squared_error: 0.0244 - acc: 0.9817 - val_loss: 0.3034 - val_mean_squared_error: 0.0936 - val_acc: 0.8650\n",
      "274/274 [==============================] - 0s 44us/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(['loss', 'mean_squared_error', 'acc'],\n",
       " [0.3033647326222301, 0.09362178442687014, 0.8649635040847055])"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Fit the model\n",
    "model.fit(x_train_norm, y_train, validation_data=(x_test_norm, y_test), epochs=3000, batch_size=512,\n",
    "          callbacks=[clr, stop], verbose=2)\n",
    "\n",
    "# evaluate the model\n",
    "scores = model.evaluate(x_test_norm, y_test)\n",
    "model.metrics_names, scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "nn_preds = model.predict_classes(x_test_norm)\n",
    "nn_pos = model.predict_proba(x_test_norm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.8613138686131386,\n",
       " 0.30866156917979454,\n",
       " 0.9448924731182796,\n",
       " 0.3071082929254639)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy(nn_preds), rmse(nn_pos), auroc(nn_pos), logloss(nn_pos)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.8461538461538461,\n",
       " 0.3220160590119148,\n",
       " 0.927314497995894,\n",
       " 0.32741931771805405)"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(0.8461538461538461,\n",
    " 0.3220160590119148,\n",
    " 0.927314497995894,\n",
    " 0.32741931771805405)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df['gbc_preds'] = gbc.predict_proba(x_train).T[1]\n",
    "test_df['gbc_preds'] = gbc.predict_proba(x_test).T[1]\n",
    "train_df['rfc_preds'] = rfc.predict_proba(x_train).T[1]\n",
    "test_df['rfc_preds'] = rfc.predict_proba(x_test).T[1]\n",
    "train_df['etc_preds'] = etc.predict_proba(x_train).T[1]\n",
    "test_df['etc_preds'] = etc.predict_proba(x_test).T[1]\n",
    "train_df['nn_preds'] = model.predict_proba(x_train_norm)\n",
    "test_df['nn_preds'] = model.predict_proba(x_test_norm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>d.gone.neg.l1</th>\n",
       "      <th>d.gone.neg.l2</th>\n",
       "      <th>d.neg.frac.l3</th>\n",
       "      <th>week</th>\n",
       "      <th>camp.length</th>\n",
       "      <th>base.poll</th>\n",
       "      <th>y2000</th>\n",
       "      <th>y2002</th>\n",
       "      <th>y2004</th>\n",
       "      <th>y2006</th>\n",
       "      <th>...</th>\n",
       "      <th>deminc</th>\n",
       "      <th>dem.contrib.l1</th>\n",
       "      <th>dem.contrib.l2</th>\n",
       "      <th>rep.contrib.l1</th>\n",
       "      <th>rep.contrib.l2</th>\n",
       "      <th>d.gone.neg</th>\n",
       "      <th>gbc_preds</th>\n",
       "      <th>rfc_preds</th>\n",
       "      <th>etc_preds</th>\n",
       "      <th>nn_preds</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-11</td>\n",
       "      <td>23</td>\n",
       "      <td>45.744681</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>38350.0</td>\n",
       "      <td>126536.00</td>\n",
       "      <td>132545.00</td>\n",
       "      <td>102559.36</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000121</td>\n",
       "      <td>0.006</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.042282</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-10</td>\n",
       "      <td>23</td>\n",
       "      <td>45.744681</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>58035.0</td>\n",
       "      <td>38350.00</td>\n",
       "      <td>522327.00</td>\n",
       "      <td>132545.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000146</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.041140</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-8</td>\n",
       "      <td>23</td>\n",
       "      <td>45.744681</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>167280.0</td>\n",
       "      <td>67068.67</td>\n",
       "      <td>177739.44</td>\n",
       "      <td>138533.00</td>\n",
       "      <td>1</td>\n",
       "      <td>0.999920</td>\n",
       "      <td>0.998</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.959737</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-7</td>\n",
       "      <td>23</td>\n",
       "      <td>45.744681</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>389947.0</td>\n",
       "      <td>167280.00</td>\n",
       "      <td>543465.00</td>\n",
       "      <td>177739.44</td>\n",
       "      <td>1</td>\n",
       "      <td>0.999957</td>\n",
       "      <td>1.000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.970611</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.071429</td>\n",
       "      <td>-6</td>\n",
       "      <td>23</td>\n",
       "      <td>45.744681</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>202734.0</td>\n",
       "      <td>389947.00</td>\n",
       "      <td>2415314.50</td>\n",
       "      <td>543465.00</td>\n",
       "      <td>1</td>\n",
       "      <td>0.999956</td>\n",
       "      <td>1.000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.980239</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 34 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   d.gone.neg.l1  d.gone.neg.l2  d.neg.frac.l3  week  camp.length  base.poll  \\\n",
       "1              0              0       0.000000   -11           23  45.744681   \n",
       "2              0              0       0.000000   -10           23  45.744681   \n",
       "3              1              0       0.000000    -8           23  45.744681   \n",
       "4              1              1       0.000000    -7           23  45.744681   \n",
       "5              1              1       0.071429    -6           23  45.744681   \n",
       "\n",
       "   y2000  y2002  y2004  y2006    ...     deminc  dem.contrib.l1  \\\n",
       "1      0      1      0      0    ...          1         38350.0   \n",
       "2      0      1      0      0    ...          1         58035.0   \n",
       "3      0      1      0      0    ...          1        167280.0   \n",
       "4      0      1      0      0    ...          1        389947.0   \n",
       "5      0      1      0      0    ...          1        202734.0   \n",
       "\n",
       "   dem.contrib.l2  rep.contrib.l1  rep.contrib.l2  d.gone.neg  gbc_preds  \\\n",
       "1       126536.00       132545.00       102559.36           0   0.000121   \n",
       "2        38350.00       522327.00       132545.00           0   0.000146   \n",
       "3        67068.67       177739.44       138533.00           1   0.999920   \n",
       "4       167280.00       543465.00       177739.44           1   0.999957   \n",
       "5       389947.00      2415314.50       543465.00           1   0.999956   \n",
       "\n",
       "   rfc_preds  etc_preds  nn_preds  \n",
       "1      0.006        0.0  0.042282  \n",
       "2      0.000        0.0  0.041140  \n",
       "3      0.998        1.0  0.959737  \n",
       "4      1.000        1.0  0.970611  \n",
       "5      1.000        1.0  0.980239  \n",
       "\n",
       "[5 rows x 34 columns]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "#train_df.to_csv(f'{PATH}train_preds.csv')\n",
    "#test_df.to_csv(f'{PATH}test_preds.csv')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}