KNN & NeuralNetwork.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import os\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "sns.set_style()\n",
    "import tabpy_client\n",
    "from sklearn import preprocessing, metrics, svm, ensemble"
   ]
  },
  {
   "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>id</th>\n",
       "      <th>gender</th>\n",
       "      <th>age</th>\n",
       "      <th>hypertension</th>\n",
       "      <th>heart_disease</th>\n",
       "      <th>ever_married</th>\n",
       "      <th>work_type</th>\n",
       "      <th>Residence_type</th>\n",
       "      <th>avg_glucose_level</th>\n",
       "      <th>bmi</th>\n",
       "      <th>smoking_status</th>\n",
       "      <th>stroke</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>9046</td>\n",
       "      <td>Male</td>\n",
       "      <td>67.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Private</td>\n",
       "      <td>Urban</td>\n",
       "      <td>228.69</td>\n",
       "      <td>36.6</td>\n",
       "      <td>formerly smoked</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>51676</td>\n",
       "      <td>Female</td>\n",
       "      <td>61.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Self-employed</td>\n",
       "      <td>Rural</td>\n",
       "      <td>202.21</td>\n",
       "      <td>NaN</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>31112</td>\n",
       "      <td>Male</td>\n",
       "      <td>80.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Private</td>\n",
       "      <td>Rural</td>\n",
       "      <td>105.92</td>\n",
       "      <td>32.5</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>60182</td>\n",
       "      <td>Female</td>\n",
       "      <td>49.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Private</td>\n",
       "      <td>Urban</td>\n",
       "      <td>171.23</td>\n",
       "      <td>34.4</td>\n",
       "      <td>smokes</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1665</td>\n",
       "      <td>Female</td>\n",
       "      <td>79.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Self-employed</td>\n",
       "      <td>Rural</td>\n",
       "      <td>174.12</td>\n",
       "      <td>24.0</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      id  gender   age  hypertension  heart_disease ever_married  \\\n",
       "0   9046    Male  67.0             0              1          Yes   \n",
       "1  51676  Female  61.0             0              0          Yes   \n",
       "2  31112    Male  80.0             0              1          Yes   \n",
       "3  60182  Female  49.0             0              0          Yes   \n",
       "4   1665  Female  79.0             1              0          Yes   \n",
       "\n",
       "       work_type Residence_type  avg_glucose_level   bmi   smoking_status  \\\n",
       "0        Private          Urban             228.69  36.6  formerly smoked   \n",
       "1  Self-employed          Rural             202.21   NaN     never smoked   \n",
       "2        Private          Rural             105.92  32.5     never smoked   \n",
       "3        Private          Urban             171.23  34.4           smokes   \n",
       "4  Self-employed          Rural             174.12  24.0     never smoked   \n",
       "\n",
       "   stroke  \n",
       "0       1  \n",
       "1       1  \n",
       "2       1  \n",
       "3       1  \n",
       "4       1  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stroke_df = pd.read_csv(os.path.join('Resources/healthcare-dataset-stroke-data.csv'))\n",
    "stroke_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>gender</th>\n",
       "      <th>age</th>\n",
       "      <th>hypertension</th>\n",
       "      <th>heart_disease</th>\n",
       "      <th>ever_married</th>\n",
       "      <th>work_type</th>\n",
       "      <th>Residence_type</th>\n",
       "      <th>avg_glucose_level</th>\n",
       "      <th>bmi</th>\n",
       "      <th>smoking_status</th>\n",
       "      <th>stroke</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Male</td>\n",
       "      <td>67.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Private</td>\n",
       "      <td>Urban</td>\n",
       "      <td>228.69</td>\n",
       "      <td>36.6</td>\n",
       "      <td>formerly smoked</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Female</td>\n",
       "      <td>61.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Self-employed</td>\n",
       "      <td>Rural</td>\n",
       "      <td>202.21</td>\n",
       "      <td>NaN</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Male</td>\n",
       "      <td>80.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Private</td>\n",
       "      <td>Rural</td>\n",
       "      <td>105.92</td>\n",
       "      <td>32.5</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Female</td>\n",
       "      <td>49.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Private</td>\n",
       "      <td>Urban</td>\n",
       "      <td>171.23</td>\n",
       "      <td>34.4</td>\n",
       "      <td>smokes</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Female</td>\n",
       "      <td>79.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Self-employed</td>\n",
       "      <td>Rural</td>\n",
       "      <td>174.12</td>\n",
       "      <td>24.0</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   gender   age  hypertension  heart_disease ever_married      work_type  \\\n",
       "0    Male  67.0             0              1          Yes        Private   \n",
       "1  Female  61.0             0              0          Yes  Self-employed   \n",
       "2    Male  80.0             0              1          Yes        Private   \n",
       "3  Female  49.0             0              0          Yes        Private   \n",
       "4  Female  79.0             1              0          Yes  Self-employed   \n",
       "\n",
       "  Residence_type  avg_glucose_level   bmi   smoking_status  stroke  \n",
       "0          Urban             228.69  36.6  formerly smoked       1  \n",
       "1          Rural             202.21   NaN     never smoked       1  \n",
       "2          Rural             105.92  32.5     never smoked       1  \n",
       "3          Urban             171.23  34.4           smokes       1  \n",
       "4          Rural             174.12  24.0     never smoked       1  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#ID column is not necessary. Drop\n",
    "stroke_df = stroke_df.drop(['id'], axis = 1)\n",
    "stroke_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "gender                 0\n",
      "age                    0\n",
      "hypertension           0\n",
      "heart_disease          0\n",
      "ever_married           0\n",
      "work_type              0\n",
      "Residence_type         0\n",
      "avg_glucose_level      0\n",
      "bmi                  201\n",
      "smoking_status         0\n",
      "stroke                 0\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "#Checking for nulls\n",
    "print(stroke_df.isna().sum())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "201 rows of NA for bmi column only. Consider Drop NA but do complete data exploration first."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 5110 entries, 0 to 5109\n",
      "Data columns (total 11 columns):\n",
      " #   Column             Non-Null Count  Dtype  \n",
      "---  ------             --------------  -----  \n",
      " 0   gender             5110 non-null   object \n",
      " 1   age                5110 non-null   float64\n",
      " 2   hypertension       5110 non-null   int64  \n",
      " 3   heart_disease      5110 non-null   int64  \n",
      " 4   ever_married       5110 non-null   object \n",
      " 5   work_type          5110 non-null   object \n",
      " 6   Residence_type     5110 non-null   object \n",
      " 7   avg_glucose_level  5110 non-null   float64\n",
      " 8   bmi                4909 non-null   float64\n",
      " 9   smoking_status     5110 non-null   object \n",
      " 10  stroke             5110 non-null   int64  \n",
      "dtypes: float64(3), int64(3), object(5)\n",
      "memory usage: 439.3+ KB\n"
     ]
    }
   ],
   "source": [
    "#Checking number of entries, data types\n",
    "stroke_df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    4612\n",
       "1     498\n",
       "Name: hypertension, dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Value counts for hypertension column to ensure only two choices\n",
    "stroke_df[\"hypertension\"].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    4834\n",
       "1     276\n",
       "Name: heart_disease, dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Value counts for heart diseease column to ensure only two choices\n",
    "stroke_df[\"heart_disease\"].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    4861\n",
       "1     249\n",
       "Name: stroke, dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Value counts for stroke column to ensure only two choices\n",
    "stroke_df[\"stroke\"].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "40"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Only 249 patients with stroke. How many will we lose if we Drop NA for the bmi column?\n",
    "bmi_null_stroke = stroke_df.loc[(stroke_df['bmi'].isna() == True) & (stroke_df['stroke'] == 1)]\n",
    "len(bmi_null_stroke)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "40 is too many rows of valuable data to lose here. Consider other options. Insert mean into BMI? Other ideas?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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></th>\n",
       "      <th>counts</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>column</th>\n",
       "      <th>value</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Residence_type</th>\n",
       "      <th>Rural</th>\n",
       "      <td>2514</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Urban</th>\n",
       "      <td>2596</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">ever_married</th>\n",
       "      <th>No</th>\n",
       "      <td>1757</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>3353</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">gender</th>\n",
       "      <th>Other</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Male</th>\n",
       "      <td>2115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Female</th>\n",
       "      <td>2994</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"4\" valign=\"top\">smoking_status</th>\n",
       "      <th>smokes</th>\n",
       "      <td>789</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>formerly smoked</th>\n",
       "      <td>885</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Unknown</th>\n",
       "      <td>1544</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>never smoked</th>\n",
       "      <td>1892</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">work_type</th>\n",
       "      <th>Never_worked</th>\n",
       "      <td>22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Govt_job</th>\n",
       "      <td>657</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>children</th>\n",
       "      <td>687</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Self-employed</th>\n",
       "      <td>819</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Private</th>\n",
       "      <td>2925</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                counts\n",
       "column         value                  \n",
       "Residence_type Rural              2514\n",
       "               Urban              2596\n",
       "ever_married   No                 1757\n",
       "               Yes                3353\n",
       "gender         Other                 1\n",
       "               Male               2115\n",
       "               Female             2994\n",
       "smoking_status smokes              789\n",
       "               formerly smoked     885\n",
       "               Unknown            1544\n",
       "               never smoked       1892\n",
       "work_type      Never_worked         22\n",
       "               Govt_job            657\n",
       "               children            687\n",
       "               Self-employed       819\n",
       "               Private            2925"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Check the value counts of all columns with datatype 'object' for a guide to encoding needs\n",
    "counts = stroke_df.select_dtypes(include=object).columns.tolist()\n",
    "(pd.DataFrame(\n",
    "    stroke_df[counts]\n",
    "    .melt(var_name='column', value_name='value')\n",
    "    .value_counts())\n",
    ".rename(columns={0: 'counts'})\n",
    ".sort_values(by=['column', 'counts']))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note that gender only has one 'other'; probably easier to remove that point for binary encoding. Note that one of the work_types is \"children,\" meaning the dataset includes children. Look at age more closely."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>gender</th>\n",
       "      <th>age</th>\n",
       "      <th>hypertension</th>\n",
       "      <th>heart_disease</th>\n",
       "      <th>ever_married</th>\n",
       "      <th>work_type</th>\n",
       "      <th>Residence_type</th>\n",
       "      <th>avg_glucose_level</th>\n",
       "      <th>bmi</th>\n",
       "      <th>smoking_status</th>\n",
       "      <th>stroke</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3295</th>\n",
       "      <td>Male</td>\n",
       "      <td>0.08</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>children</td>\n",
       "      <td>Rural</td>\n",
       "      <td>70.33</td>\n",
       "      <td>16.9</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1614</th>\n",
       "      <td>Female</td>\n",
       "      <td>0.08</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>children</td>\n",
       "      <td>Urban</td>\n",
       "      <td>139.67</td>\n",
       "      <td>14.1</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3618</th>\n",
       "      <td>Male</td>\n",
       "      <td>0.16</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>children</td>\n",
       "      <td>Urban</td>\n",
       "      <td>114.71</td>\n",
       "      <td>17.4</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4021</th>\n",
       "      <td>Male</td>\n",
       "      <td>0.16</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>children</td>\n",
       "      <td>Urban</td>\n",
       "      <td>109.52</td>\n",
       "      <td>13.9</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3968</th>\n",
       "      <td>Male</td>\n",
       "      <td>0.16</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>children</td>\n",
       "      <td>Rural</td>\n",
       "      <td>69.79</td>\n",
       "      <td>13.0</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      gender   age  hypertension  heart_disease ever_married work_type  \\\n",
       "3295    Male  0.08             0              0           No  children   \n",
       "1614  Female  0.08             0              0           No  children   \n",
       "3618    Male  0.16             0              0           No  children   \n",
       "4021    Male  0.16             0              0           No  children   \n",
       "3968    Male  0.16             0              0           No  children   \n",
       "\n",
       "     Residence_type  avg_glucose_level   bmi smoking_status  stroke  \n",
       "3295          Rural              70.33  16.9        Unknown       0  \n",
       "1614          Urban             139.67  14.1        Unknown       0  \n",
       "3618          Urban             114.71  17.4        Unknown       0  \n",
       "4021          Urban             109.52  13.9        Unknown       0  \n",
       "3968          Rural              69.79  13.0        Unknown       0  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Examining age column\n",
    "stroke_df.sort_values(by=['age']).head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Dropped the \"Other\" in the \"Gender\" column"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This dataset even contains babies. Children are likely less useful in predicting stroke, although these particular children may have some other health issues, which is why they were included. How many children actually are labeled with stroke?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Children stroke count\n",
    "children_stroke = stroke_df.loc[(stroke_df['age'] < 17) & (stroke_df['stroke'] == 1)]\n",
    "len(children_stroke)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Remove children from the dataset in this instance and compare model performance with the children data left in vs deleted."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>gender</th>\n",
       "      <th>age</th>\n",
       "      <th>hypertension</th>\n",
       "      <th>heart_disease</th>\n",
       "      <th>ever_married</th>\n",
       "      <th>work_type</th>\n",
       "      <th>Residence_type</th>\n",
       "      <th>avg_glucose_level</th>\n",
       "      <th>bmi</th>\n",
       "      <th>smoking_status</th>\n",
       "      <th>stroke</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>395</th>\n",
       "      <td>Female</td>\n",
       "      <td>17.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>Private</td>\n",
       "      <td>Urban</td>\n",
       "      <td>87.52</td>\n",
       "      <td>39.2</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3777</th>\n",
       "      <td>Female</td>\n",
       "      <td>17.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>Private</td>\n",
       "      <td>Rural</td>\n",
       "      <td>68.66</td>\n",
       "      <td>35.1</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4611</th>\n",
       "      <td>Female</td>\n",
       "      <td>17.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>Never_worked</td>\n",
       "      <td>Urban</td>\n",
       "      <td>78.08</td>\n",
       "      <td>44.9</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3256</th>\n",
       "      <td>Female</td>\n",
       "      <td>17.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>Private</td>\n",
       "      <td>Urban</td>\n",
       "      <td>89.58</td>\n",
       "      <td>22.8</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2128</th>\n",
       "      <td>Male</td>\n",
       "      <td>17.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>Private</td>\n",
       "      <td>Rural</td>\n",
       "      <td>61.67</td>\n",
       "      <td>97.6</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4452</th>\n",
       "      <td>Female</td>\n",
       "      <td>82.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Private</td>\n",
       "      <td>Urban</td>\n",
       "      <td>115.71</td>\n",
       "      <td>31.1</td>\n",
       "      <td>formerly smoked</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>Male</td>\n",
       "      <td>82.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Private</td>\n",
       "      <td>Urban</td>\n",
       "      <td>144.90</td>\n",
       "      <td>26.4</td>\n",
       "      <td>smokes</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2314</th>\n",
       "      <td>Male</td>\n",
       "      <td>82.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Private</td>\n",
       "      <td>Urban</td>\n",
       "      <td>89.83</td>\n",
       "      <td>24.7</td>\n",
       "      <td>smokes</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>Female</td>\n",
       "      <td>82.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Self-employed</td>\n",
       "      <td>Urban</td>\n",
       "      <td>196.92</td>\n",
       "      <td>22.2</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5062</th>\n",
       "      <td>Female</td>\n",
       "      <td>82.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Self-employed</td>\n",
       "      <td>Rural</td>\n",
       "      <td>211.58</td>\n",
       "      <td>36.9</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4314 rows × 11 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      gender   age  hypertension  heart_disease ever_married      work_type  \\\n",
       "395   Female  17.0             0              0           No        Private   \n",
       "3777  Female  17.0             0              0           No        Private   \n",
       "4611  Female  17.0             0              0           No   Never_worked   \n",
       "3256  Female  17.0             0              0           No        Private   \n",
       "2128    Male  17.0             1              0           No        Private   \n",
       "...      ...   ...           ...            ...          ...            ...   \n",
       "4452  Female  82.0             1              0          Yes        Private   \n",
       "42      Male  82.0             0              1          Yes        Private   \n",
       "2314    Male  82.0             0              0          Yes        Private   \n",
       "32    Female  82.0             1              0          Yes  Self-employed   \n",
       "5062  Female  82.0             0              0          Yes  Self-employed   \n",
       "\n",
       "     Residence_type  avg_glucose_level   bmi   smoking_status  stroke  \n",
       "395           Urban              87.52  39.2     never smoked       0  \n",
       "3777          Rural              68.66  35.1     never smoked       0  \n",
       "4611          Urban              78.08  44.9     never smoked       0  \n",
       "3256          Urban              89.58  22.8          Unknown       0  \n",
       "2128          Rural              61.67  97.6          Unknown       0  \n",
       "...             ...                ...   ...              ...     ...  \n",
       "4452          Urban             115.71  31.1  formerly smoked       0  \n",
       "42            Urban             144.90  26.4           smokes       1  \n",
       "2314          Urban              89.83  24.7           smokes       0  \n",
       "32            Urban             196.92  22.2     never smoked       1  \n",
       "5062          Rural             211.58  36.9     never smoked       0  \n",
       "\n",
       "[4314 rows x 11 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Removing children from the dataset\n",
    "\n",
    "no_children_df = stroke_df[stroke_df.age > 16]\n",
    "no_children_df.sort_values(by=['age'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Private          2840\n",
       "Self-employed     810\n",
       "Govt_job          654\n",
       "Never_worked       10\n",
       "Name: work_type, dtype: int64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Ensure children is no longer listed as a work type\n",
    "no_children_df.sort_values(by=['age'])\n",
    "no_children_df['work_type'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>gender</th>\n",
       "      <th>age</th>\n",
       "      <th>hypertension</th>\n",
       "      <th>heart_disease</th>\n",
       "      <th>ever_married</th>\n",
       "      <th>work_type</th>\n",
       "      <th>Residence_type</th>\n",
       "      <th>avg_glucose_level</th>\n",
       "      <th>bmi</th>\n",
       "      <th>smoking_status</th>\n",
       "      <th>stroke</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Male</td>\n",
       "      <td>67.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Private</td>\n",
       "      <td>Urban</td>\n",
       "      <td>228.69</td>\n",
       "      <td>36.6</td>\n",
       "      <td>formerly smoked</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Female</td>\n",
       "      <td>61.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Self-employed</td>\n",
       "      <td>Rural</td>\n",
       "      <td>202.21</td>\n",
       "      <td>NaN</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Male</td>\n",
       "      <td>80.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Private</td>\n",
       "      <td>Rural</td>\n",
       "      <td>105.92</td>\n",
       "      <td>32.5</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Female</td>\n",
       "      <td>49.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Private</td>\n",
       "      <td>Urban</td>\n",
       "      <td>171.23</td>\n",
       "      <td>34.4</td>\n",
       "      <td>smokes</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Female</td>\n",
       "      <td>79.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Self-employed</td>\n",
       "      <td>Rural</td>\n",
       "      <td>174.12</td>\n",
       "      <td>24.0</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5105</th>\n",
       "      <td>Female</td>\n",
       "      <td>80.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Private</td>\n",
       "      <td>Urban</td>\n",
       "      <td>83.75</td>\n",
       "      <td>NaN</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5106</th>\n",
       "      <td>Female</td>\n",
       "      <td>81.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Self-employed</td>\n",
       "      <td>Urban</td>\n",
       "      <td>125.20</td>\n",
       "      <td>40.0</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5107</th>\n",
       "      <td>Female</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Self-employed</td>\n",
       "      <td>Rural</td>\n",
       "      <td>82.99</td>\n",
       "      <td>30.6</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5108</th>\n",
       "      <td>Male</td>\n",
       "      <td>51.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Private</td>\n",
       "      <td>Rural</td>\n",
       "      <td>166.29</td>\n",
       "      <td>25.6</td>\n",
       "      <td>formerly smoked</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5109</th>\n",
       "      <td>Female</td>\n",
       "      <td>44.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Govt_job</td>\n",
       "      <td>Urban</td>\n",
       "      <td>85.28</td>\n",
       "      <td>26.2</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4313 rows × 11 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      gender   age  hypertension  heart_disease ever_married      work_type  \\\n",
       "0       Male  67.0             0              1          Yes        Private   \n",
       "1     Female  61.0             0              0          Yes  Self-employed   \n",
       "2       Male  80.0             0              1          Yes        Private   \n",
       "3     Female  49.0             0              0          Yes        Private   \n",
       "4     Female  79.0             1              0          Yes  Self-employed   \n",
       "...      ...   ...           ...            ...          ...            ...   \n",
       "5105  Female  80.0             1              0          Yes        Private   \n",
       "5106  Female  81.0             0              0          Yes  Self-employed   \n",
       "5107  Female  35.0             0              0          Yes  Self-employed   \n",
       "5108    Male  51.0             0              0          Yes        Private   \n",
       "5109  Female  44.0             0              0          Yes       Govt_job   \n",
       "\n",
       "     Residence_type  avg_glucose_level   bmi   smoking_status  stroke  \n",
       "0             Urban             228.69  36.6  formerly smoked       1  \n",
       "1             Rural             202.21   NaN     never smoked       1  \n",
       "2             Rural             105.92  32.5     never smoked       1  \n",
       "3             Urban             171.23  34.4           smokes       1  \n",
       "4             Rural             174.12  24.0     never smoked       1  \n",
       "...             ...                ...   ...              ...     ...  \n",
       "5105          Urban              83.75   NaN     never smoked       0  \n",
       "5106          Urban             125.20  40.0     never smoked       0  \n",
       "5107          Rural              82.99  30.6     never smoked       0  \n",
       "5108          Rural             166.29  25.6  formerly smoked       0  \n",
       "5109          Urban              85.28  26.2          Unknown       0  \n",
       "\n",
       "[4313 rows x 11 columns]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Drop the 'other' row in gender column\n",
    "cleaned_stroke_df = no_children_df[no_children_df.gender != 'Other']\n",
    "cleaned_stroke_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>age</th>\n",
       "      <th>hypertension</th>\n",
       "      <th>heart_disease</th>\n",
       "      <th>work_type</th>\n",
       "      <th>avg_glucose_level</th>\n",
       "      <th>bmi</th>\n",
       "      <th>smoking_status</th>\n",
       "      <th>stroke</th>\n",
       "      <th>gender_Female</th>\n",
       "      <th>gender_Male</th>\n",
       "      <th>ever_married_No</th>\n",
       "      <th>ever_married_Yes</th>\n",
       "      <th>Residence_type_Rural</th>\n",
       "      <th>Residence_type_Urban</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>67.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Private</td>\n",
       "      <td>228.69</td>\n",
       "      <td>36.6</td>\n",
       "      <td>formerly smoked</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>61.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Self-employed</td>\n",
       "      <td>202.21</td>\n",
       "      <td>NaN</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>80.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Private</td>\n",
       "      <td>105.92</td>\n",
       "      <td>32.5</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>49.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Private</td>\n",
       "      <td>171.23</td>\n",
       "      <td>34.4</td>\n",
       "      <td>smokes</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>79.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>Self-employed</td>\n",
       "      <td>174.12</td>\n",
       "      <td>24.0</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5105</th>\n",
       "      <td>80.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>Private</td>\n",
       "      <td>83.75</td>\n",
       "      <td>NaN</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5106</th>\n",
       "      <td>81.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Self-employed</td>\n",
       "      <td>125.20</td>\n",
       "      <td>40.0</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5107</th>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Self-employed</td>\n",
       "      <td>82.99</td>\n",
       "      <td>30.6</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5108</th>\n",
       "      <td>51.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Private</td>\n",
       "      <td>166.29</td>\n",
       "      <td>25.6</td>\n",
       "      <td>formerly smoked</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5109</th>\n",
       "      <td>44.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Govt_job</td>\n",
       "      <td>85.28</td>\n",
       "      <td>26.2</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4313 rows × 14 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       age  hypertension  heart_disease      work_type  avg_glucose_level  \\\n",
       "0     67.0             0              1        Private             228.69   \n",
       "1     61.0             0              0  Self-employed             202.21   \n",
       "2     80.0             0              1        Private             105.92   \n",
       "3     49.0             0              0        Private             171.23   \n",
       "4     79.0             1              0  Self-employed             174.12   \n",
       "...    ...           ...            ...            ...                ...   \n",
       "5105  80.0             1              0        Private              83.75   \n",
       "5106  81.0             0              0  Self-employed             125.20   \n",
       "5107  35.0             0              0  Self-employed              82.99   \n",
       "5108  51.0             0              0        Private             166.29   \n",
       "5109  44.0             0              0       Govt_job              85.28   \n",
       "\n",
       "       bmi   smoking_status  stroke  gender_Female  gender_Male  \\\n",
       "0     36.6  formerly smoked       1              0            1   \n",
       "1      NaN     never smoked       1              1            0   \n",
       "2     32.5     never smoked       1              0            1   \n",
       "3     34.4           smokes       1              1            0   \n",
       "4     24.0     never smoked       1              1            0   \n",
       "...    ...              ...     ...            ...          ...   \n",
       "5105   NaN     never smoked       0              1            0   \n",
       "5106  40.0     never smoked       0              1            0   \n",
       "5107  30.6     never smoked       0              1            0   \n",
       "5108  25.6  formerly smoked       0              0            1   \n",
       "5109  26.2          Unknown       0              1            0   \n",
       "\n",
       "      ever_married_No  ever_married_Yes  Residence_type_Rural  \\\n",
       "0                   0                 1                     0   \n",
       "1                   0                 1                     1   \n",
       "2                   0                 1                     1   \n",
       "3                   0                 1                     0   \n",
       "4                   0                 1                     1   \n",
       "...               ...               ...                   ...   \n",
       "5105                0                 1                     0   \n",
       "5106                0                 1                     0   \n",
       "5107                0                 1                     1   \n",
       "5108                0                 1                     1   \n",
       "5109                0                 1                     0   \n",
       "\n",
       "      Residence_type_Urban  \n",
       "0                        1  \n",
       "1                        0  \n",
       "2                        0  \n",
       "3                        1  \n",
       "4                        0  \n",
       "...                    ...  \n",
       "5105                     1  \n",
       "5106                     1  \n",
       "5107                     0  \n",
       "5108                     0  \n",
       "5109                     1  \n",
       "\n",
       "[4313 rows x 14 columns]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Use Pandas get_dummies to convert categorical data\n",
    "\n",
    "encoded_df = pd.get_dummies(cleaned_stroke_df, columns=['gender','ever_married', 'Residence_type'])\n",
    "encoded_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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>age</th>\n",
       "      <th>hypertension</th>\n",
       "      <th>heart_disease</th>\n",
       "      <th>work_type</th>\n",
       "      <th>avg_glucose_level</th>\n",
       "      <th>bmi</th>\n",
       "      <th>smoking_status</th>\n",
       "      <th>stroke</th>\n",
       "      <th>gender_Male</th>\n",
       "      <th>ever_married_Yes</th>\n",
       "      <th>Residence_type_Urban</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>67.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Private</td>\n",
       "      <td>228.69</td>\n",
       "      <td>36.6</td>\n",
       "      <td>formerly smoked</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>61.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Self-employed</td>\n",
       "      <td>202.21</td>\n",
       "      <td>NaN</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>80.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Private</td>\n",
       "      <td>105.92</td>\n",
       "      <td>32.5</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>49.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Private</td>\n",
       "      <td>171.23</td>\n",
       "      <td>34.4</td>\n",
       "      <td>smokes</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>79.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>Self-employed</td>\n",
       "      <td>174.12</td>\n",
       "      <td>24.0</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5105</th>\n",
       "      <td>80.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>Private</td>\n",
       "      <td>83.75</td>\n",
       "      <td>NaN</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5106</th>\n",
       "      <td>81.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Self-employed</td>\n",
       "      <td>125.20</td>\n",
       "      <td>40.0</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5107</th>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Self-employed</td>\n",
       "      <td>82.99</td>\n",
       "      <td>30.6</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5108</th>\n",
       "      <td>51.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Private</td>\n",
       "      <td>166.29</td>\n",
       "      <td>25.6</td>\n",
       "      <td>formerly smoked</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5109</th>\n",
       "      <td>44.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Govt_job</td>\n",
       "      <td>85.28</td>\n",
       "      <td>26.2</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4313 rows × 11 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       age  hypertension  heart_disease      work_type  avg_glucose_level  \\\n",
       "0     67.0             0              1        Private             228.69   \n",
       "1     61.0             0              0  Self-employed             202.21   \n",
       "2     80.0             0              1        Private             105.92   \n",
       "3     49.0             0              0        Private             171.23   \n",
       "4     79.0             1              0  Self-employed             174.12   \n",
       "...    ...           ...            ...            ...                ...   \n",
       "5105  80.0             1              0        Private              83.75   \n",
       "5106  81.0             0              0  Self-employed             125.20   \n",
       "5107  35.0             0              0  Self-employed              82.99   \n",
       "5108  51.0             0              0        Private             166.29   \n",
       "5109  44.0             0              0       Govt_job              85.28   \n",
       "\n",
       "       bmi   smoking_status  stroke  gender_Male  ever_married_Yes  \\\n",
       "0     36.6  formerly smoked       1            1                 1   \n",
       "1      NaN     never smoked       1            0                 1   \n",
       "2     32.5     never smoked       1            1                 1   \n",
       "3     34.4           smokes       1            0                 1   \n",
       "4     24.0     never smoked       1            0                 1   \n",
       "...    ...              ...     ...          ...               ...   \n",
       "5105   NaN     never smoked       0            0                 1   \n",
       "5106  40.0     never smoked       0            0                 1   \n",
       "5107  30.6     never smoked       0            0                 1   \n",
       "5108  25.6  formerly smoked       0            1                 1   \n",
       "5109  26.2          Unknown       0            0                 1   \n",
       "\n",
       "      Residence_type_Urban  \n",
       "0                        1  \n",
       "1                        0  \n",
       "2                        0  \n",
       "3                        1  \n",
       "4                        0  \n",
       "...                    ...  \n",
       "5105                     1  \n",
       "5106                     1  \n",
       "5107                     0  \n",
       "5108                     0  \n",
       "5109                     1  \n",
       "\n",
       "[4313 rows x 11 columns]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Drop the extra dummies columns\n",
    "encoded_stroke_df = encoded_df.drop(columns=['gender_Female','ever_married_No','Residence_type_Rural'])\n",
    "encoded_stroke_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "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>age</th>\n",
       "      <th>hypertension</th>\n",
       "      <th>heart_disease</th>\n",
       "      <th>work_type</th>\n",
       "      <th>avg_glucose_level</th>\n",
       "      <th>bmi</th>\n",
       "      <th>smoking_status</th>\n",
       "      <th>stroke</th>\n",
       "      <th>gender_Male</th>\n",
       "      <th>ever_married_Yes</th>\n",
       "      <th>Residence_type_Urban</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>67.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>228.69</td>\n",
       "      <td>36.6</td>\n",
       "      <td>formerly smoked</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>61.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>202.21</td>\n",
       "      <td>NaN</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>80.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>105.92</td>\n",
       "      <td>32.5</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>49.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>171.23</td>\n",
       "      <td>34.4</td>\n",
       "      <td>smokes</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>79.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>174.12</td>\n",
       "      <td>24.0</td>\n",
       "      <td>never smoked</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    age  hypertension  heart_disease  work_type  avg_glucose_level   bmi  \\\n",
       "0  67.0             0              1          2             228.69  36.6   \n",
       "1  61.0             0              0          3             202.21   NaN   \n",
       "2  80.0             0              1          2             105.92  32.5   \n",
       "3  49.0             0              0          2             171.23  34.4   \n",
       "4  79.0             1              0          3             174.12  24.0   \n",
       "\n",
       "    smoking_status  stroke  gender_Male  ever_married_Yes  \\\n",
       "0  formerly smoked       1            1                 1   \n",
       "1     never smoked       1            0                 1   \n",
       "2     never smoked       1            1                 1   \n",
       "3           smokes       1            0                 1   \n",
       "4     never smoked       1            0                 1   \n",
       "\n",
       "   Residence_type_Urban  \n",
       "0                     1  \n",
       "1                     0  \n",
       "2                     0  \n",
       "3                     1  \n",
       "4                     0  "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Label encode the categorical columns that have more than two choices\n",
    "# label_encoder object knows how to understand word labels. \n",
    "label_encoder = preprocessing.LabelEncoder()\n",
    "# Encode labels in column 'Country'. \n",
    "encoded_stroke_df['work_type']= label_encoder.fit_transform(encoded_stroke_df['work_type']) \n",
    "encoded_stroke_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "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>age</th>\n",
       "      <th>hypertension</th>\n",
       "      <th>heart_disease</th>\n",
       "      <th>work_type</th>\n",
       "      <th>avg_glucose_level</th>\n",
       "      <th>bmi</th>\n",
       "      <th>smoking_status</th>\n",
       "      <th>stroke</th>\n",
       "      <th>gender_Male</th>\n",
       "      <th>ever_married_Yes</th>\n",
       "      <th>Residence_type_Urban</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>67.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>228.69</td>\n",
       "      <td>36.6</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>61.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>202.21</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>80.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>105.92</td>\n",
       "      <td>32.5</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>49.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>171.23</td>\n",
       "      <td>34.4</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>79.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>174.12</td>\n",
       "      <td>24.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    age  hypertension  heart_disease  work_type  avg_glucose_level   bmi  \\\n",
       "0  67.0             0              1          2             228.69  36.6   \n",
       "1  61.0             0              0          3             202.21   NaN   \n",
       "2  80.0             0              1          2             105.92  32.5   \n",
       "3  49.0             0              0          2             171.23  34.4   \n",
       "4  79.0             1              0          3             174.12  24.0   \n",
       "\n",
       "   smoking_status  stroke  gender_Male  ever_married_Yes  Residence_type_Urban  \n",
       "0               1       1            1                 1                     1  \n",
       "1               2       1            0                 1                     0  \n",
       "2               2       1            1                 1                     0  \n",
       "3               3       1            0                 1                     1  \n",
       "4               2       1            0                 1                     0  "
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# label_encoder object knows how to understand word labels. \n",
    "label_encoder = preprocessing.LabelEncoder()\n",
    "# Encode labels in column 'Country'. \n",
    "encoded_stroke_df['smoking_status']= label_encoder.fit_transform(encoded_stroke_df['smoking_status']) \n",
    "encoded_stroke_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[<AxesSubplot:title={'center':'bmi'}>]], dtype=object)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Histogram of bmi column to view outliers\n",
    "encoded_stroke_df.hist(column = 'bmi')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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>age</th>\n",
       "      <th>hypertension</th>\n",
       "      <th>heart_disease</th>\n",
       "      <th>work_type</th>\n",
       "      <th>avg_glucose_level</th>\n",
       "      <th>bmi</th>\n",
       "      <th>smoking_status</th>\n",
       "      <th>stroke</th>\n",
       "      <th>gender_Male</th>\n",
       "      <th>ever_married_Yes</th>\n",
       "      <th>Residence_type_Urban</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2128</th>\n",
       "      <td>17.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>61.67</td>\n",
       "      <td>97.6</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4209</th>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>56.90</td>\n",
       "      <td>92.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>928</th>\n",
       "      <td>23.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>70.03</td>\n",
       "      <td>78.0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>544</th>\n",
       "      <td>42.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>210.48</td>\n",
       "      <td>71.9</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1559</th>\n",
       "      <td>53.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>72.63</td>\n",
       "      <td>66.8</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       age  hypertension  heart_disease  work_type  avg_glucose_level   bmi  \\\n",
       "2128  17.0             1              0          2              61.67  97.6   \n",
       "4209  38.0             1              0          2              56.90  92.0   \n",
       "928   23.0             1              0          2              70.03  78.0   \n",
       "544   42.0             0              0          2             210.48  71.9   \n",
       "1559  53.0             0              0          2              72.63  66.8   \n",
       "\n",
       "      smoking_status  stroke  gender_Male  ever_married_Yes  \\\n",
       "2128               0       0            1                 0   \n",
       "4209               2       0            1                 1   \n",
       "928                3       0            0                 0   \n",
       "544                2       0            1                 1   \n",
       "1559               0       0            0                 1   \n",
       "\n",
       "      Residence_type_Urban  \n",
       "2128                     0  \n",
       "4209                     0  \n",
       "928                      1  \n",
       "544                      0  \n",
       "1559                     0  "
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "encoded_stroke_df.sort_values(by=['bmi'], ascending = False).head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "12.349999999999994\n",
      "47.150000000000006\n",
      "117\n"
     ]
    }
   ],
   "source": [
    "#Determine outliers using upper and lower bounds\n",
    "quartiles = encoded_stroke_df['bmi'].quantile([.25,.5,.75])\n",
    "lowerq = quartiles[0.25]\n",
    "upperq = quartiles[0.75]\n",
    "iqr = upperq-lowerq\n",
    "lower_bound = lowerq - (1.5*iqr)\n",
    "upper_bound = upperq + (1.5*iqr)\n",
    "outliers = encoded_stroke_df.loc[(encoded_stroke_df['bmi'] < lower_bound) | (encoded_stroke_df['bmi'] > upper_bound)]\n",
    "        \n",
    "#Display upper & lower bounds and number of outliers\n",
    "       \n",
    "print(lower_bound)\n",
    "print(upper_bound)\n",
    "print(len(outliers))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "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>age</th>\n",
       "      <th>hypertension</th>\n",
       "      <th>heart_disease</th>\n",
       "      <th>work_type</th>\n",
       "      <th>avg_glucose_level</th>\n",
       "      <th>bmi</th>\n",
       "      <th>smoking_status</th>\n",
       "      <th>stroke</th>\n",
       "      <th>gender_Male</th>\n",
       "      <th>ever_married_Yes</th>\n",
       "      <th>Residence_type_Urban</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>4313.000000</td>\n",
       "      <td>4313.000000</td>\n",
       "      <td>4313.000000</td>\n",
       "      <td>4313.000000</td>\n",
       "      <td>4313.000000</td>\n",
       "      <td>4130.000000</td>\n",
       "      <td>4313.000000</td>\n",
       "      <td>4313.000000</td>\n",
       "      <td>4313.000000</td>\n",
       "      <td>4313.000000</td>\n",
       "      <td>4313.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>49.745885</td>\n",
       "      <td>0.115465</td>\n",
       "      <td>0.063761</td>\n",
       "      <td>1.882217</td>\n",
       "      <td>108.226388</td>\n",
       "      <td>30.421719</td>\n",
       "      <td>1.570369</td>\n",
       "      <td>0.057269</td>\n",
       "      <td>0.394157</td>\n",
       "      <td>0.777417</td>\n",
       "      <td>0.50939</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>18.125129</td>\n",
       "      <td>0.319619</td>\n",
       "      <td>0.244355</td>\n",
       "      <td>0.884856</td>\n",
       "      <td>47.568050</td>\n",
       "      <td>7.323256</td>\n",
       "      <td>1.010025</td>\n",
       "      <td>0.232382</td>\n",
       "      <td>0.488726</td>\n",
       "      <td>0.416029</td>\n",
       "      <td>0.49997</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>17.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>55.120000</td>\n",
       "      <td>11.300000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>35.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>77.450000</td>\n",
       "      <td>25.400000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>50.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>92.270000</td>\n",
       "      <td>29.200000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>64.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>115.920000</td>\n",
       "      <td>34.100000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>82.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>271.740000</td>\n",
       "      <td>97.600000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.00000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               age  hypertension  heart_disease    work_type  \\\n",
       "count  4313.000000   4313.000000    4313.000000  4313.000000   \n",
       "mean     49.745885      0.115465       0.063761     1.882217   \n",
       "std      18.125129      0.319619       0.244355     0.884856   \n",
       "min      17.000000      0.000000       0.000000     0.000000   \n",
       "25%      35.000000      0.000000       0.000000     2.000000   \n",
       "50%      50.000000      0.000000       0.000000     2.000000   \n",
       "75%      64.000000      0.000000       0.000000     2.000000   \n",
       "max      82.000000      1.000000       1.000000     3.000000   \n",
       "\n",
       "       avg_glucose_level          bmi  smoking_status       stroke  \\\n",
       "count        4313.000000  4130.000000     4313.000000  4313.000000   \n",
       "mean          108.226388    30.421719        1.570369     0.057269   \n",
       "std            47.568050     7.323256        1.010025     0.232382   \n",
       "min            55.120000    11.300000        0.000000     0.000000   \n",
       "25%            77.450000    25.400000        1.000000     0.000000   \n",
       "50%            92.270000    29.200000        2.000000     0.000000   \n",
       "75%           115.920000    34.100000        2.000000     0.000000   \n",
       "max           271.740000    97.600000        3.000000     1.000000   \n",
       "\n",
       "       gender_Male  ever_married_Yes  Residence_type_Urban  \n",
       "count  4313.000000       4313.000000            4313.00000  \n",
       "mean      0.394157          0.777417               0.50939  \n",
       "std       0.488726          0.416029               0.49997  \n",
       "min       0.000000          0.000000               0.00000  \n",
       "25%       0.000000          1.000000               0.00000  \n",
       "50%       0.000000          1.000000               1.00000  \n",
       "75%       1.000000          1.000000               1.00000  \n",
       "max       1.000000          1.000000               1.00000  "
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#quick look at the data\n",
    "encoded_stroke_df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "age                     50.00\n",
       "hypertension             0.00\n",
       "heart_disease            0.00\n",
       "work_type                2.00\n",
       "avg_glucose_level       92.27\n",
       "bmi                     29.20\n",
       "smoking_status           2.00\n",
       "stroke                   0.00\n",
       "gender_Male              0.00\n",
       "ever_married_Yes         1.00\n",
       "Residence_type_Urban     1.00\n",
       "dtype: float64"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Get the median values to compare mean and median in bmi\n",
    "encoded_stroke_df.median()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Mean and median are nearly identical. Either is an acceptable replacement."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "30.421719128329308"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Set mean to a variable\n",
    "mean_bmi = encoded_stroke_df['bmi'].mean()\n",
    "mean_bmi"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "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>age</th>\n",
       "      <th>hypertension</th>\n",
       "      <th>heart_disease</th>\n",
       "      <th>work_type</th>\n",
       "      <th>avg_glucose_level</th>\n",
       "      <th>bmi</th>\n",
       "      <th>smoking_status</th>\n",
       "      <th>stroke</th>\n",
       "      <th>gender_Male</th>\n",
       "      <th>ever_married_Yes</th>\n",
       "      <th>Residence_type_Urban</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5105</th>\n",
       "      <td>80.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>83.75</td>\n",
       "      <td>30.421719</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5106</th>\n",
       "      <td>81.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>125.20</td>\n",
       "      <td>40.000000</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5107</th>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>82.99</td>\n",
       "      <td>30.600000</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5108</th>\n",
       "      <td>51.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>166.29</td>\n",
       "      <td>25.600000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5109</th>\n",
       "      <td>44.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>85.28</td>\n",
       "      <td>26.200000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       age  hypertension  heart_disease  work_type  avg_glucose_level  \\\n",
       "5105  80.0             1              0          2              83.75   \n",
       "5106  81.0             0              0          3             125.20   \n",
       "5107  35.0             0              0          3              82.99   \n",
       "5108  51.0             0              0          2             166.29   \n",
       "5109  44.0             0              0          0              85.28   \n",
       "\n",
       "            bmi  smoking_status  stroke  gender_Male  ever_married_Yes  \\\n",
       "5105  30.421719               2       0            0                 1   \n",
       "5106  40.000000               2       0            0                 1   \n",
       "5107  30.600000               2       0            0                 1   \n",
       "5108  25.600000               1       0            1                 1   \n",
       "5109  26.200000               0       0            0                 1   \n",
       "\n",
       "      Residence_type_Urban  \n",
       "5105                     1  \n",
       "5106                     1  \n",
       "5107                     0  \n",
       "5108                     0  \n",
       "5109                     1  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#replace NaN with mean bmi\n",
    "replace_stroke_df = encoded_stroke_df.fillna(mean_bmi)\n",
    "stroke_df = replace_stroke_df\n",
    "stroke_df.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Model work starts here"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Feature Engineering\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.model_selection import GridSearchCV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "y = stroke_df[\"stroke\"]\n",
    "# target_names = [\"negative\", \"positive\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "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>age</th>\n",
       "      <th>hypertension</th>\n",
       "      <th>heart_disease</th>\n",
       "      <th>work_type</th>\n",
       "      <th>avg_glucose_level</th>\n",
       "      <th>bmi</th>\n",
       "      <th>smoking_status</th>\n",
       "      <th>gender_Male</th>\n",
       "      <th>ever_married_Yes</th>\n",
       "      <th>Residence_type_Urban</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>67.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>228.69</td>\n",
       "      <td>36.600000</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>61.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>202.21</td>\n",
       "      <td>30.421719</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>80.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>105.92</td>\n",
       "      <td>32.500000</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>49.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>171.23</td>\n",
       "      <td>34.400000</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>79.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>174.12</td>\n",
       "      <td>24.000000</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    age  hypertension  heart_disease  work_type  avg_glucose_level        bmi  \\\n",
       "0  67.0             0              1          2             228.69  36.600000   \n",
       "1  61.0             0              0          3             202.21  30.421719   \n",
       "2  80.0             0              1          2             105.92  32.500000   \n",
       "3  49.0             0              0          2             171.23  34.400000   \n",
       "4  79.0             1              0          3             174.12  24.000000   \n",
       "\n",
       "   smoking_status  gender_Male  ever_married_Yes  Residence_type_Urban  \n",
       "0               1            1                 1                     1  \n",
       "1               2            0                 1                     0  \n",
       "2               2            1                 1                     0  \n",
       "3               3            0                 1                     1  \n",
       "4               2            0                 1                     0  "
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X = stroke_df.drop(\"stroke\", axis=1)\n",
    "X.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from matplotlib import style\n",
    "style.use(\"ggplot\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# from sklearn.datasets.samples_generator import make_blobs\n",
    "# X, y = make_blobs(n_samples=100, centers=2, random_state=0, cluster_std=.95)\n",
    "plt.scatter(X.iloc[:, 0], X.iloc[:, 1], c=y, s=100, cmap=\"bwr\");\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0       67.0\n",
      "1       61.0\n",
      "2       80.0\n",
      "3       49.0\n",
      "4       79.0\n",
      "        ... \n",
      "5105    80.0\n",
      "5106    81.0\n",
      "5107    35.0\n",
      "5108    51.0\n",
      "5109    44.0\n",
      "Name: age, Length: 4313, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "print(X.iloc[:, 0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "# Create a StandardScater model and fit it to the training data\n",
    "\n",
    "X_scaler = StandardScaler().fit(X_train)\n",
    "\n",
    "# Transform the training and testing data using the X_scaler and y_scaler models\n",
    "\n",
    "X_train_scaled = X_scaler.transform(X_train)\n",
    "X_test_scaled = X_scaler.transform(X_test)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.neighbors import KNeighborsClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "KNeighborsClassifier()"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "knn_model = KNeighborsClassifier()\n",
    "knn_model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is the KNN model to find the Nearest Neighbors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "run_control": {
     "marked": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "k: 1, Train/Test Score: 1.000/0.896\n",
      "k: 3, Train/Test Score: 0.952/0.927\n",
      "k: 5, Train/Test Score: 0.947/0.931\n",
      "k: 7, Train/Test Score: 0.947/0.933\n",
      "k: 9, Train/Test Score: 0.947/0.933\n",
      "k: 11, Train/Test Score: 0.946/0.933\n",
      "k: 13, Train/Test Score: 0.946/0.933\n",
      "k: 15, Train/Test Score: 0.946/0.934\n",
      "k: 17, Train/Test Score: 0.946/0.934\n",
      "k: 19, Train/Test Score: 0.946/0.934\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "knn_train_scores = []\n",
    "knn_test_scores = []\n",
    "for k in range(1, 20, 2):\n",
    "    knn_model = KNeighborsClassifier(n_neighbors=k)\n",
    "    knn_model.fit(X_train_scaled, y_train)\n",
    "    knn_train_score = knn_model.score(X_train_scaled, y_train)\n",
    "    knn_test_score = knn_model.score(X_test_scaled, y_test)\n",
    "    knn_train_scores.append(knn_train_score)\n",
    "    knn_test_scores.append(knn_test_score)\n",
    "    print(f\"k: {k}, Train/Test Score: {knn_train_score:.3f}/{knn_test_score:.3f}\")\n",
    "    \n",
    "plt.plot(range(1, 20, 2), knn_train_scores, marker='o')\n",
    "plt.plot(range(1, 20, 2), knn_test_scores, marker=\"x\")\n",
    "plt.xlabel(\"k neighbors\")\n",
    "plt.ylabel(\"Testing accuracy Score\")\n",
    "plt.title(\"KNN Neighbors\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create the GridSearch estimator along with a parameter object containing the values to adjust\n",
    "knn_param_grid = {'n_neighbors': range(1, 20, 2)}\n",
    "knn_grid = GridSearchCV(knn_model, knn_param_grid, verbose=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 10 candidates, totalling 50 fits\n",
      "[CV] n_neighbors=1 ...................................................\n",
      "[CV] ....................... n_neighbors=1, score=0.892, total=   0.0s\n",
      "[CV] n_neighbors=1 ...................................................\n",
      "[CV] ....................... n_neighbors=1, score=0.901, total=   0.0s\n",
      "[CV] n_neighbors=1 ...................................................\n",
      "[CV] ....................... n_neighbors=1, score=0.915, total=   0.0s\n",
      "[CV] n_neighbors=1 ...................................................\n",
      "[CV] ....................... n_neighbors=1, score=0.904, total=   0.0s\n",
      "[CV] n_neighbors=1 ...................................................\n",
      "[CV] ....................... n_neighbors=1, score=0.898, total=   0.0s\n",
      "[CV] n_neighbors=3 ...................................................\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.0s remaining:    0.0s\n",
      "[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:    0.0s remaining:    0.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV] ....................... n_neighbors=3, score=0.932, total=   0.0s\n",
      "[CV] n_neighbors=3 ...................................................\n",
      "[CV] ....................... n_neighbors=3, score=0.938, total=   0.0s\n",
      "[CV] n_neighbors=3 ...................................................\n",
      "[CV] ....................... n_neighbors=3, score=0.943, total=   0.0s\n",
      "[CV] n_neighbors=3 ...................................................\n",
      "[CV] ....................... n_neighbors=3, score=0.934, total=   0.0s\n",
      "[CV] n_neighbors=3 ...................................................\n",
      "[CV] ....................... n_neighbors=3, score=0.937, total=   0.0s\n",
      "[CV] n_neighbors=5 ...................................................\n",
      "[CV] ....................... n_neighbors=5, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=5 ...................................................\n",
      "[CV] ....................... n_neighbors=5, score=0.943, total=   0.0s\n",
      "[CV] n_neighbors=5 ...................................................\n",
      "[CV] ....................... n_neighbors=5, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=5 ...................................................\n",
      "[CV] ....................... n_neighbors=5, score=0.944, total=   0.0s\n",
      "[CV] n_neighbors=5 ...................................................\n",
      "[CV] ....................... n_neighbors=5, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=7 ...................................................\n",
      "[CV] ....................... n_neighbors=7, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=7 ...................................................\n",
      "[CV] ....................... n_neighbors=7, score=0.944, total=   0.0s\n",
      "[CV] n_neighbors=7 ...................................................\n",
      "[CV] ....................... n_neighbors=7, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=7 ...................................................\n",
      "[CV] ....................... n_neighbors=7, score=0.944, total=   0.0s\n",
      "[CV] n_neighbors=7 ...................................................\n",
      "[CV] ....................... n_neighbors=7, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=9 ...................................................\n",
      "[CV] ....................... n_neighbors=9, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=9 ...................................................\n",
      "[CV] ....................... n_neighbors=9, score=0.944, total=   0.0s\n",
      "[CV] n_neighbors=9 ...................................................\n",
      "[CV] ....................... n_neighbors=9, score=0.946, total=   0.1s\n",
      "[CV] n_neighbors=9 ...................................................\n",
      "[CV] ....................... n_neighbors=9, score=0.944, total=   0.1s\n",
      "[CV] n_neighbors=9 ...................................................\n",
      "[CV] ....................... n_neighbors=9, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=11 ..................................................\n",
      "[CV] ...................... n_neighbors=11, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=11 ..................................................\n",
      "[CV] ...................... n_neighbors=11, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=11 ..................................................\n",
      "[CV] ...................... n_neighbors=11, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=11 ..................................................\n",
      "[CV] ...................... n_neighbors=11, score=0.944, total=   0.0s\n",
      "[CV] n_neighbors=11 ..................................................\n",
      "[CV] ...................... n_neighbors=11, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=13 ..................................................\n",
      "[CV] ...................... n_neighbors=13, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=13 ..................................................\n",
      "[CV] ...................... n_neighbors=13, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=13 ..................................................\n",
      "[CV] ...................... n_neighbors=13, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=13 ..................................................\n",
      "[CV] ...................... n_neighbors=13, score=0.944, total=   0.0s\n",
      "[CV] n_neighbors=13 ..................................................\n",
      "[CV] ...................... n_neighbors=13, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=15 ..................................................\n",
      "[CV] ...................... n_neighbors=15, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=15 ..................................................\n",
      "[CV] ...................... n_neighbors=15, score=0.946, total=   0.1s\n",
      "[CV] n_neighbors=15 ..................................................\n",
      "[CV] ...................... n_neighbors=15, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=15 ..................................................\n",
      "[CV] ...................... n_neighbors=15, score=0.944, total=   0.0s\n",
      "[CV] n_neighbors=15 ..................................................\n",
      "[CV] ...................... n_neighbors=15, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=17 ..................................................\n",
      "[CV] ...................... n_neighbors=17, score=0.946, total=   0.1s\n",
      "[CV] n_neighbors=17 ..................................................\n",
      "[CV] ...................... n_neighbors=17, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=17 ..................................................\n",
      "[CV] ...................... n_neighbors=17, score=0.946, total=   0.1s\n",
      "[CV] n_neighbors=17 ..................................................\n",
      "[CV] ...................... n_neighbors=17, score=0.944, total=   0.0s\n",
      "[CV] n_neighbors=17 ..................................................\n",
      "[CV] ...................... n_neighbors=17, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=19 ..................................................\n",
      "[CV] ...................... n_neighbors=19, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=19 ..................................................\n",
      "[CV] ...................... n_neighbors=19, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=19 ..................................................\n",
      "[CV] ...................... n_neighbors=19, score=0.946, total=   0.0s\n",
      "[CV] n_neighbors=19 ..................................................\n",
      "[CV] ...................... n_neighbors=19, score=0.944, total=   0.0s\n",
      "[CV] n_neighbors=19 ..................................................\n",
      "[CV] ...................... n_neighbors=19, score=0.946, total=   0.0s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done  50 out of  50 | elapsed:    2.2s finished\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(estimator=KNeighborsClassifier(n_neighbors=19),\n",
       "             param_grid={'n_neighbors': range(1, 20, 2)}, verbose=3)"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "knn_grid.fit(X_train_scaled, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'n_neighbors': 11}"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "knn_grid.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 0, ..., 0, 0, 0], dtype=int64)"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Make predictions with the hypertuned model\n",
    "predictions = knn_grid.predict(X_test_scaled)\n",
    "predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1007    1]\n",
      " [  71    0]]\n"
     ]
    }
   ],
   "source": [
    "#Confusion matrix to see correct and incorrect predictions\n",
    "from sklearn.metrics import confusion_matrix\n",
    "confusion_matrix = confusion_matrix(y_test, predictions)\n",
    "print(confusion_matrix)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-44-473d05fc1822>:16: RuntimeWarning: invalid value encountered in double_scalars\n",
      "  f1_score_stroke  = 2*precision_stroke*recall_stroke / (precision_stroke + recall_stroke)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'KNN Neighbor')"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYMAAAFsCAYAAAAudtVFAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAAB2UUlEQVR4nO3dd1gU19fA8S9LR6SD2BGxxh5b7EaMJcaYoib2FnuJLbHEqLGbWKPGrrH3XqLBjhV7Q7GhYqMKIp2d9w9e5udKW0BlNeeTZ5/I7JQ7u7Nz5t65c66RoigKQggh/tM0OV0AIYQQOU+CgRBCCAkGQgghJBgIIYRAgoEQQggkGAghhECCgfiPc3Nz448//tB7fn9/f4yMjDh79mya8xw+fBgjIyOCg4PfRBGFeCckGAi9dOrUiWbNmulM27VrF1ZWVowcORKAMWPGYGRkRLdu3XTmS+0EamRkhJmZGXfv3s1wO6/LzrKv8/HxoXfv3plaRogPkQQDkSUrV67km2++YdKkSUyYMEGdbmFhwfLly7l27VqG6zA2NlYDSWZlZ9lXOTs7Y2Vlle31vAtxcXE5XQTxAZNgIDJt1qxZdOvWjcWLFzNgwACd94oWLUqjRo0YPnx4huvp168fGzZs4Ny5c5kug77LLlu2jNKlS2NhYUHx4sWZMWMGWq1Wff/1ZiI/Pz/q1q2LhYUFJUqUYM+ePVhbW7N8+XKd9d6/f5+GDRtiZWVF6dKl+ffff1Ns+9SpU1SoUAELCws+/vjjFGXdsmULZcuWxdzcnIIFCzJhwgReTQjg5ubGmDFj6NKlC3Z2drRt2zYzH5EQmSLBQGTKqFGjGD58OFu2bKF9+/apzjN58mR2797NsWPH0l1X1apV+eabb/jpp58yXQ59ll20aBEjRozgt99+w9fXl2nTpjFlyhTmzZuX6vxarZavvvoKExMTTp06xfLlyxk7diyxsbEp5h05ciT9+/fn0qVLVKlShe+++47IyEideYYMGcKUKVM4e/Ys7u7ufP7550RFRQFw7tw5WrZsyddff82VK1eYPHkykyZNYs6cOTrrmD59OiVLluTs2bNMnDgxsx+TEPpThNBDx44dFTMzMwVQdu3aleo8o0ePVj766CNFURSlU6dOSvXq1RVFUZR79+4pgOLj46POCygbN25Ubt26pZiamip79+5Vt/P555+nWxZ9ly1YsKCyYsUKnWVnzJihlCpVSv27cOHCyu+//64oiqL8888/irGxsRIQEKC+f/z4cQVQli1bprMv8+fPV+cJCAhQAOXYsWOKoijKoUOHFEBZtWqVOs+LFy8UW1tbZdGiRYqiKEqbNm2U+vXrp/j88ufPr1O2Zs2apftZCPGmSM1A6K1MmTIULVqUsWPH8vz583Tn/e2337h48SJbtmxJdz4PDw9++OEHhg0bptN8o4/0lg0KCuLhw4f06NEDa2tr9TVs2DDu3LmT6vpu3LhBvnz5yJ8/vzqtSpUqaDQpfyblypVT/50vXz4AAgMDdeb55JNP1H9bW1tTtmxZrl+/DoCvry81a9bUmb9WrVo8evSIiIgIdVrlypXT/QyEeFMkGAi95c2blyNHjhAeHo6npydhYWFpzluwYEH69evH8OHDSUhISHe9o0eP5s6dO6xevTrTZUpr2eTgMH/+fC5evKi+rl69mubNbUVRMDIy0mu7pqam6r+Tl8lMMEtvW69Oz5Url97rFCI7JBiITMmfPz+HDx/m5cuXNGjQgJCQkDTnHT58OEFBQSxevDjddbq4uDBkyBBGjRqVavt8VpbNkycP+fPn586dO3h4eKR4paZUqVI8evSIx48fq9POnj2b6RpLslOnTqn/fvnyJVevXqVUqVIAlC5dGm9vb535vb29KVCgALlz587S9oTIDgkGItPy5s3L4cOHiYuL49NPP03z4Sp7e3tGjBjBrFmzMlzn4MGDiYmJYdu2bZkuT1rLjhkzhqlTpzJjxgxu3rzJ1atXWbFiBZMmTUp1PQ0bNqREiRJ07NiRS5cucerUKQYNGoSJiYneNYZXjR8/nn///Zdr167RpUsXzMzMaNOmjVrmI0eOMGbMGPz8/Fi9ejXTpk3L0s10Id4ECQYiS/LkycOhQ4cAqF+/for28mT9+vXDxcUlw/VZW1szevRoYmJiMl2WtJbt1q0bS5cuZeXKlZQvX57atWuzcOFCihQpkup6NBoNW7duJTY2lqpVq9KxY0dGjhyJkZERFhYWmS7X5MmTGTx4MJUqVeLWrVvs2rVLbfapVKkSGzduZPPmzZQpU4Zhw4YxbNgw+vbtm+ntCPEmGCmKjHQmRFouXbpEhQoVOHv2LB9//HFOF0eIt0aCgRCv2Lp1K7ly5aJYsWL4+/szaNAgFEXhwoULWWoqEuJ9YZLTBRDCkLx48YKff/6Zhw8fYm9vT7169ZgxY4YEAvHBk5qBEEIIqRkIIYShmDdvHufPn8fW1pZp06YBEBkZyYwZMwgKCsLZ2ZmBAwdibW0NJDVrHjx4EI1GQ+fOnalQoQIAd+/eZe7cucTFxVGxYkU6d+6cYe32gwwG8cF3M55J/KdY5qud00UQBioh7lG2ls/M+cbUyT3d9+vVq0fjxo2ZO3euOm3btm2ULVuWFi1asG3bNrZt20a7du0ICAjgxIkTTJ8+nbCwMMaNG8esWbPQaDQsWrSIHj16UKxYMSZNmsTFixepWLFiutuWrqVCCGEgSpcurV71J/Px8aFu3boA1K1bFx8fH3V6jRo1MDU1xcXFBVdXV27fvk1YWBjR0dEUL14cIyMj6tSpoy6Tng+yZiCEEO+MNlHvWb28vPDy8lL/9vT0xNPTM91lwsPDsbe3B5Ie5EzOXRUaGkqxYsXU+RwcHAgNDcXY2BhHR0d1uqOjI6GhoRmWTYKBEEJkR2L6ubdepc/JX19p9f3Jap8gaSYSQohsUBSt3q+ssLW1VZNChoWFYWNjAyRd8b+aGyw0NBQHB4cU00NCQnBwcMhwOxIMhBAiO7Ra/V9ZULlyZY4cOQLAkSNHqFKlijr9xIkTxMfHExgYyJMnT/Dw8MDe3h5LS0v8/PxQFIWjR4/qlQr9g3zOQHoTiddJbyKRluz2Jop7eEnvec0Klk/3/ZkzZ3L9+nVevHiBra0trVq1okqVKsyYMYPg4GCcnJwYNGiQepN5y5YtHDp0CI1GQ6dOndQeQ3fu3GHevHnExcVRoUIFunTpkmHXUgkG4j9BgoFIS7aDwf3zes9rVrhStrb1NskNZCGEyI4s3gswNBIMhBAiG5RM9CYyZBIMhBAiO7J4Y9jQSDAQQojskGYiIYQQmXkC2ZBJMBBCiOyQmoEQQojMpKMwZBIMhBAiO+QGshBCCEWRewZCCCHknoEQQghpJhJCCCE1AyGEEEBifE6X4I2QYCCEENkhzURCCCGkmUgIIYTUDIQQQiDBQAghBChyA1kIIYTcMxBCCCHNREIIIZCagRBCCKRmIIQQAqkZCCGEABJkcBshhBBSMxBCCCH3DIQQQkjNQAghBFIzEEIIgdQMhBBCIL2JhBBCAIqS0yV4IyQYGIjn4RF07T8cgODQMIw1GuztbAFYt3gmpqam2d5Gp74/ERUVw4alswG46uvHH3MXs3zO1GyvW7wdsdEPuHL1hvr3N9924f79gFTnfR7qh51D8Wxtb8niGdSpXZ3wiBdotVr69x/JqdPnsrXOD57cMxBvkp2tDZv/ngvA3CWrsLK0oHObb9X3ExISMTExzvZ2Qp8/59hJH2p/UiXb6xJvX3R0DJWrfPZOt/nz8PFs2bKbhp51mDdvMpU+bvhOt//ekWAg3raR46dha5MbX787lC7hgZWVpU6QaNGuJ3N/H0v+vHnYue8gqzduJz4+gXIfleCXwX0wNk4ZPDp//w0L/l6XIhgkJiYy469l+Fy4TFx8PN9//QWtWjRFq9UyYfo8zl64Qv58rihaLV81+4zP6td+J5+B0JUrlxVbNy/Dzt4WU1MTfh09lZ079+vM4+rqwtrVf5HbJjcmJsb07Tsc7+NnaOhZh9G/DsHM3Iy7d+/TtdtAXr6MSnNbR4+dxqNoEQB+HNCdTp1aA7B06Vpm/7kYKytL1q1ZQP4CeTE21jBh4iw2btzx9nbeUMkNZPEu+D98xOJZEzE2NmbuklWpznPH/wH/HDjCyvnTMDUxYdwfc9i1/xBfNvFMMW/5MqU4cPQkZ85dwsrKUp2+Zdc+clvnYv2S2cTFxdGu5xBqVK3E9Zu3ePzkGVtX/kVo2HOat+3BV83e7ZXqf5mlpQVnfZJO9v7+D2j9XQ++admVFy8icXS05/ixnSmCwffffcX+f48wafJsNBoNVlaWODraM2L4AD5r3JqoqGiGDunNwB+7M37CzDS33axZQ65e9aVSxbJ07NiKGjWbYWRkxInjuzh67CRFihTm8ZOnNG/RAQAbm9xv7XMwaImJb3R1u3bt4uDBgxgZGVGwYEF69+5NXFwcM2bMICgoCGdnZwYOHIi1tTUAW7du5eDBg2g0Gjp37kyFChWytF0JBgauUf1aqV7hv+r02Ytcv3Gb77oOACA2NhYHe7s05+/R6TsW/L2Wgb26qNNOnDmP3x1/9h/yBiDy5UvuP3zE+UvX+OzT2mg0GpwcHahSsVz2d0ro7fVmIhMTE8aPG0bt2tXQahXy53clTx5nnj0LUuc5e/YiixZOw9TUhO079nHp0jXq1vmEUqWKc/TIdgDMzEw5dSr1ewFTJv3CiOEDCA4K4YceQ/i0fi22b/+HqKhoALZt20utmtXYt/8wUyePYtLEEeze7YX38TNv8ZMwYG+wmSg0NJS9e/cyY8YMzMzMmD59OidOnCAgIICyZcvSokULtm3bxrZt22jXrh0BAQGcOHGC6dOnExYWxrhx45g1axYajSbT25ZgYOAsLS3Uf5sYG6O80nMhNi4OAEVRaN7Ek4G9Ouu1zmofV+DPRSu5fO1/NyYVBUYM7EXNah/rzHv0pE92ii/esDbff42zsyNVqzUhISGB236nsLAw15nnmPdp6jf4hqZNGrB82SymTZ/P87BwvA4cpV37PhluI/meQbIGn6beJHjr1l2qVm9CkyafMmH8cP71OpJuTeOD9YbvGWi1WuLi4jA2NiYuLg57e3u2bt3KmDFjAKhbty5jxoyhXbt2+Pj4UKNGDUxNTXFxccHV1ZXbt29TvHjmOxJkPnyIHJMvbx6u37wNwPWbt3n05BkA1StX4N/D3oSEPQcgPOIFj58+S3dd3Tu0ZumaTerfNatVYv3W3cT/f59p/wcBREXHUKncR/x7+DharZbg0DB8Llx+C3sm9GVrm5vAwGASEhKoV7cGbm4FU8xTqFB+AgODWbJ0DcuWraNihbKcOn2OGp9UoWhRNyDpIqNYMXe9tnns2CmaN2+EpaUFVlaWfPllY7yPnyZv3jxERUWzZs0Wps2YT8WKZd/krr4/FK3eLy8vL4YNG6a+vLy8dFbl4ODAF198Qa9evejevTtWVlaUL1+e8PBw7O3tAbC3tyciIgJIqkk4OjrqLB8aGpql3ZCawXukYb2a7NjrxTcd+1CmVHEKF8wPQNEihen3Qwe6/zgSraLF1MSEkYN6k881T5rrqlOjKg7/33UV4JsvGvPoSSCtOvdDURTs7WyZPflXGtaryamzF2nRviduBQtQ7qOSWOfK9db3VaRuzdotbN/6N6dO7uHSpWv43riVYp66dWsweFBP4uMTeBn5kk5dBhAcHErXbgNZtXIu5uZmAPw6eiq3bt3NcJsXLl5lxYqNnDyRVFtYunQtFy9e47OGdZk8+Re0WoX4+Hj69h3+Znf2PaFo9X/OwNPTE0/PlPfykkVGRuLj48PcuXOxsrJi+vTpHD16NO1tv8FnHIyUN7k2AxEfnPEBLvQXFRWNlZUlz8Mj+K7bAFbNn4aTo0NOFytTLPNJ7yeRuoS4R9laPmr+AL3nteo5K933T548ycWLF+nVqxcAR44c4datW1y5coUxY8Zgb29PWFgYY8aMYdasWWzduhWAr776CoAJEybQsmXLLDUTSc1AZKj3T6N58eIl8Qnx9OzU5r0LBEK8VW+wN5GTkxO3bt0iNjYWMzMzrly5QtGiRTE3N+fIkSO0aNGCI0eOUKVKUtfwypUrM3v2bJo1a0ZYWBhPnjzBw8MjS9uWYCAyJE8oC5GON3gDuVixYlSvXp2ff/4ZY2Nj3Nzc8PT0JCYmhhkzZnDw4EGcnJwYNGgQAAULFuSTTz5h0KBBaDQaunbtmqWeRCDNROI/QpqJRFqy3Uw0q6fe81oNmJ+tbb1NUjN4z/wycTpHj5/Bwd6ObauSDqzwiBcMHjWJx0+fkc81D9PGDcf2/x8AWrRiPVt27cNYo2H4/3cdffkyig69h6rrfBYUTLPP6jPsR/0PavF+WrRwGp839SQwKJgKFRvkdHE+DB/I9bR0LX3PtGjakPnTx+tMW7xyA9UrV2DP+iVUr1yBJas2AHDn3n32HjjC9lXzmT99POP+mENiYiK5clmx+e+56iufqwue9WrmxO6Id2zFig183qxtThfjw6LV6v8yYAYZDBRF4ejRo2zalNQPPjg4mNu3b+dwqQxD5Qpl1av+ZIeOnVRTT3zZxJODR08CcPDYKZo0qIuZmRkF8rlSqEA+rvj66Sx7/+EjQsKe83H5Mu9mB0SOOuZ9mtD/fx5FvCFaRf+XATPIYLB48WL8/Pw4fvw4ABYWFixZsiSHS2W4QsKe4+yU1MPH2cmB0OfhAAQGheCax1mdL4+LE4FBwTrL7vn3MI0b1MHIyOjdFViID0liov4vA2aQweD27dt069ZNzeFvbW1NQgajCb36ZJ9IopDySsQI3ZP+3gNHaOpZ7x2VSIgPj6LV6v0yZAZ5A9nY2BitVqterUZERGR45frqk33/td5EjvZ2BAWH4uzkQFBwqPpkcR5nJ56+ksDsWWAwzs7/e3T9xq27JCZq+ahksXdeZiE+GAbe/KMvg6wZNGnShN9//53w8HDWrl3LqFGj1CfsREr1alVn+96kHCfb93pRv/YnANSvVZ29B44QFxdHwOOnPAh4TNlS/3syca/XYZp41s2RMgvxwchEbiJDZpA1g+rVq+Pu7s6VK1cAGDp0KHZ2djlbKAMxdPRkfC5c5vnzCBq0aEfvru3p1r4Vg0dNZMuufeTN48z08SMB8HAvTKNPa9O8bQ9MjI0ZOai3TjrsfQePMe+P33JqV0QOWLVyLnXrfIKTkwP+d88y9rc/WLZ8XU4X6/32gdQMDPKhs0mTJjF06FBMTJJiVVhYGJMnT2bKlCl6Lf9fayYSGZOHzkRasvvQ2ctfv9N73ly/GW7gNchmoipVqjB9+nS0Wi2BgYFMmDCBNm3a5HSxhBAiJWkmens8PT1JSEhg6tSpBAUF0b17d0qUKJHTxRJCiJQ+kGYig6oZ7Nq1S33Fx8cTEhKCm5sbt27dYteuXTldvBy1csM2WrTryZdte7ByfVLa2j8XruCrDr34pmMffvhxBIFBISmWi42N47tuA/i6Y2++bNuDOYtXqu9Nn7eErzr0Yvi4P9RpO/45wMoN2976/og3r9Fn9bh29Sg3rnvz09CUI5qVKFEU76M7ePniLoMG9lCnOzk5cOTQVi5eOEDz5o3U6Vs2LyVv3rTHxBBJPpSupQYVDKKjo9VXTEwMVatWxdXVVZ32X3Xrrj+bd/zD2sUz2fz3PI6cOMP9h4/o3PYbtq74i81/z6VuzWr8tWxNimXNzExZOnsyW/6ex6a/53L89DkuXfXlReRLLl7xZeuKv9AmJuJ35x4xsbFs3/Mv333dLAf2UmSHRqNh9qwJNPuiHWXL16d16xaUKqXbZTg09Dk/DhzF9BkLdKZ/17oFK1ZupFbt5gwZlJRHv9nnDblw4QpPnqQ/Yp7gg3kC2aCaiVq2bKnzd3R0NEZGRlhYWKSxxH/DXf+HlPuoJJb//zlUrlCWA0dP0KXt/z6v6OgYUnsUw8jICCsrSwASEhJISEjAyMgIjZER8QkJKIpCTGwcJiYmLFu9ibYtv8TUxKAOC6GHqlUqcueOP/fuPQBgw4btNP+iEb6+/xsJLSgohKCgEJo21U1QFx+fgKWlBebmZmi1WoyNjenfrxtfftXxne7De8vAT/L6Mshf/YMHD5gzZw6RkZEA5M6dm759+1KwYMrxXv8LPNwLM3vh3zwPj8Dc3IxjJ33UB8VmLVjOjn8OkDtXLpb+OTnV5RMTE2nVpT8PHj3m+6+bUe6jkkDSMJrfdupL9coVyJ0rF1dv+NGriyQxex/ly+/Kw4DH6t8Bj55QtUpFvZZdu24rq1bMpV27bxkxYiK9enZk5epNREfHvK3iflgMPM2EvgwyGCxcuJAOHTpQpkxS8rRr166xYMECxo8fn8GSH6aiboXo0rYlP/w4AitLS4p7uKvPCwzo0YkBPTqxaMV61mzeSd9u7VMsb2xszOa/5xLxIpIBw8dx664/xdzd6NK2pVq7+HXSTPp2a8+mHf9w0uc8xYsWoUen79/pfoqsS+0JfX17jUdEvKB5iw4A2NnZMnRIb75t1Y35f03F3t6OGTMWcOr0uTda3g9JZsZANmQGdc8gWWxsrBoIAD766CNiY2NzsEQ575svGrFx2Rz+nvc7tja5KVwwv877n39WD6/Dx9Ndh01ua6pUKof3qbM60339kjLCFi5YgJ3/HGDauBHcuuvP/YfZ638t3p1HAU8oWCCf+neB/Hmz1N4/auRAJk2ezXetW3D+/BW6/TCI8eMk31e6PpB7BgYZDFxcXNi0aROBgYEEBgayefNmnJ2dM17wAxby/2mHnzwN5MCR4zTxrKtzsj507BRFChdIsVxo2HMiXiQ1t8XExnLK5wJFCus2t/25aCV9u7UnISGBxP/v8aDRaIiO+W8H4PeJz9mLeHgUwc2tIKamprRq9SU7d+3P1Do8PIqQN18ejh47hZWVJVqtFkVRsLAwf0ul/kB8IOMZGGQzUa9evdiwYQPTpk0DoFSpUvTu3TuHS5WzBo4Yz/OICExMTBg5uDe2NrkZPXkW/g8CMNIYkc/VhV+H9gOSUlePnjyTv6aNIygkjJHj/yBRq0XRKjT6tDb1alZT13vg6AnKlCqOy/8nsCtfpiRfte9F8aJulCzmniP7KjIvMTGRAT/+wp7dazDWaFj+93quX/ej+w9JzYYLF60kTx5nTp/ci42NNVqtlv79fqBs+Xq8+P+LhXG//cyoX5Oe8l+3fhtbNi2lX7+ujBn7R5rbFRj8Fb++DC4dhVarZcKECYwaNSrL65B0FOJ1ko5CpCW76She9Gys97y55/+TrW29TQZXM9BoNJiZmREVFYWVlVVOF0cIIdKlJBp284++DC4YAJiamjJ48GDKlSuHufn/2iu7dOmSg6USQohUfCDNRAYZDCpVqkSlSpV0psmwjEIIQ/ShdC01yGAQFRVF06ZNdabt2bMnh0ojhBDp+ECCgUF2LT1y5EiKaYcPH373BRFCiIxoM/EyYAZVM/D29sbb25vAwECdgWyio6PJnTt3DpZMCCFSpyQY+FleTwYVDEqUKIG9vT0vXrzgiy++UKdbWFhQuHDhHCyZEEKk4cOIBYYVDJydnXF2dmbChAkAvHjxAl9fXywsLHTG7hVCCEPxodxANqh7BpMnT+bBg6QUvGFhYQwePJiDBw/y559/snv37hwunRBCpELuGbx5gYGBFCpUCIBDhw5Rrlw5+vbtS3R0NKNGjeLzzz/P4RIKIYQuqRm8Ba82BV29epWKFZPysVtaWspzBkIIwyQ1gzfP0dGRvXv34ujoyL1796hQoQIAcXFxJH4gA0gIIT4sSkJOl+DNMKiaQa9evXj48CGHDx/mxx9/JFeuXAD4+flRr169nC2cEEKkQtHq/zJkBpe19E2QrKXidZK1VKQlu1lLgxvV1Xtep30pH6g1FAbVTCSEEO+bN33F//LlS+bPn8/Dhw8xMjKiV69e5MuXjxkzZhAUFISzszMDBw7E2toagK1bt3Lw4EE0Gg2dO3dWm9czS4KBEEJkw5sOBsuWLaNChQoMHjyYhIQEYmNj2bp1K2XLlqVFixZs27aNbdu20a5dOwICAjhx4gTTp08nLCyMcePGMWvWLDSazN8BMKh7BkII8b5REo30fmUkKioKX19fPv30UwBMTEzIlSsXPj4+1K2b1BxVt25dfHx8APDx8aFGjRqYmpri4uKCq6srt2/fztJ+GGTNICQkhKVLl3Ljxg00Gg0lSpSgc+fOODo65nTRhBBCR2ZqBl5eXnh5eal/e3p64unpqf4dGBiIjY0N8+bN4/79+7i7u9OpUyfCw8Oxt7cHwN7enoiICABCQ0MpVqyYuryDgwOhoaFZ2g+DDAbz5s2jVq1aDBo0CIBjx44xb968bA2FKYQQb4Oi1f8ZqNdP/q9LTEzk3r17dOnShWLFirFs2TK2bduW9rbfYP8fg2wmioiIoH79+hgbG2NsbEy9evXUSCiEEIbkTXYtdXR0xNHRUb3ar169Ovfu3cPW1pawsDAgKVWPjY2NOn9ISIi6fGhoKA4ODlnaD4MMBjY2Nhw9ehStVotWq+Xo0aOSwloIYZAUxUjvV0bs7OxwdHTk8ePHAFy5coUCBQpQuXJldZyXI0eOUKVKFQAqV67MiRMniI+PJzAwkCdPnuDh4ZGl/TDI5wyCg4NZsmQJfn5+GBkZUbx4cTp37oyzs7Ney8tzBuJ18pyBSEt2nzMIqPap3vMWOH0ww3n8/f2ZP38+CQkJuLi40Lt3bxRFYcaMGQQHB+Pk5MSgQYPUrqVbtmzh0KFDaDQaOnXqpKbxySyDDAbZJcFAvE6CgUhLdoPBg8oN9J630NkD2drW22RQN5A3bdqU7vvffvvtOyqJEELoJzM3kA2ZQQUDc3PzFNNiY2M5ePAgL168kGAghDA4EgzegleHuoyOjmbPnj0cOnSIGjVq6LwnhBCG4kNpaDeoYAAQGRnJrl27OHbsGHXr1mXKlCnqjRIhhDA0UjN4C1auXMmZM2do0KAB06ZNw8LCIqeLJIQQ6dKny+j7wKCCwa5duzAxMWHLli1s3bpVna4oCkZGRvz99985WDohhEgpUY+cQ+8DgwoG69evz+kiCCFEpkjNQAghhNwzEEIIIb2JhBBCIDUDIYQQQKLWIPN9ZpoEAyGEyAZpJhJCCIFWehMJIYSQrqVCCCGkmciQFSnePKeLIIT4j5BmIiGEENKbSAghBHwgrUQSDIQQIjukmUgIIYT0JhJCCAHanC7AGyLBQAghskFBagZCCPGflyDNREIIIaRmIIQQQu4ZCCGEkJqBEEIIpGYghBACSJSagRBCiA9k1EsJBkIIkR1aqRkIIYSQRHVCCCHkBrIQQgjQGkkzkRBC/Ocl5nQB3hAJBkIIkQ1vozeRVqtl2LBhODg4MGzYMCIjI5kxYwZBQUE4OzszcOBArK2tAdi6dSsHDx5Eo9HQuXNnKlSokKVtfhjjtQkhRA7RYqT3S1979uwhf/786t/btm2jbNmyzJ49m7Jly7Jt2zYAAgICOHHiBNOnT2fkyJEsWbIErTZrdzEkGAghRDYomXjpIyQkhPPnz9OgQQN1mo+PD3Xr1gWgbt26+Pj4qNNr1KiBqakpLi4uuLq6cvv27SzthzQTCSFENmSmmcjLywsvLy/1b09PTzw9PXXmWb58Oe3atSM6OlqdFh4ejr29PQD29vZEREQAEBoaSrFixdT5HBwcCA0NzcpuSDAQQojsyEyjTGon/1edO3cOW1tb3N3duXbtWobrU5Q395SDBAMhhMiGxDd4A/nmzZucPXuWCxcuEBcXR3R0NLNnz8bW1pawsDDs7e0JCwvDxsYGAEdHR0JCQtTlQ0NDcXBwyNK25Z6BEEJkgzYTr4y0adOG+fPnM3fuXH788UfKlClD//79qVy5MkeOHAHgyJEjVKlSBYDKlStz4sQJ4uPjCQwM5MmTJ3h4eGRpP6RmIIQQ2fAunkBu0aIFM2bM4ODBgzg5OTFo0CAAChYsyCeffMKgQYPQaDR07doVjSZr1/hGyptsdDIQBRzK5HQRhIF5GhmW00UQBioh7lG2lp9fsJ3e8/Z8uCpb23qbpGYghBDZILmJhBBCSDoKIYQQMriNEEIIpJlICCEEEgyEEEIgI50JIYRA7hkIIYRAehMJIYQAtB9IQ5EEAyGEyAa5gSyEEOIDqRdIMDAI94MuceP6LfXvru36E/Dwcarz3nxwhhKFqmZre9PnjKdOvU+oUakxcXHx2DvYsefgej6p0Chb6xVvh4ODPfv3rQfANY8ziYmJBAUnDWDySY3PiY+Pz/Y2Dvy7Ede8eYiJieVl5Eu6dR+Mn9+dbK/3v0BqBuKNiYmOpVHdb9/pNhO1Wlq3/ZqVy9a/0+2KzAsNDaNylc8A+HXUICIjXzJ9xgL1fWNjYxITs38bs0OHvpw7f5luXdsyZfIvfPV152yv878gwejDqBvIeAYGyCqXJeu2LmbvoQ14eW/hsyb1U8zjkseJTbuWs+/IJryOb6Vq9UoA1Klfg+37VrH30AbmL5uGVS7LVLexeP5KfujVHmNj4xTv9ezXmV1e6/j32BYGD+ujTh8wpAeHT+1gzZZFzFk0lR59O72ZHRaZtmTxDP6YOhqv/RuZPHEkv44axKCBPdT3L144QOHCBQBo0+ZrTh7fxVmf/cybOyXDFMfHvE9RtKgbAFMm/cLFCwe4cN6Lli2bA+Dq6sKhA5s567OfixcOUKtm9mqq77s3PQZyTpGagQGwsDRn35FNADy8/4genQfRrcMAIl+8xN7Bjp3717B/7yGdZVp8+zlHDp7gz+kL0Wg0WFpZYO9gR//B3fnuqx+Ijoqmd/8udO/dkZm/z0+xzccBTzhz+gLftP6Cf/85rE6vU78GRdwL0czzO4yMjFi2Zg7VPvmY6OgYmn7RkMb1WmJsYsw/hzZy5dL1t/q5iPQVK+bOZ41bo9Vq+XXUoFTnKVnSg1Ytm1O7bgsSEhL4c/ZE2rT5mlWrNqW53mafN+Tq1Rt89VVTypf/iEofN8TJyYFTJ/Zw7Ngpvv/uK/b/e4RJk2ej0Wiwskr9guO/QpqJxBvzejORiYkJw34ZQLUaldFqtbjmdcHZxZGgwP8Nb3fp/FWm/TkOU1MT/tl9gOtXb1L9syoUL1GUbXtXAmBqZsp5n0tpbnfO9IUsXTOHA/uPqtPq1K9Bnfo11OCUK5cVRYoWJpe1Ffv3HiImJhaAf/cdfpMfgciCTVt2odWmfyr6tH4tKlUsy6mTewCwtLQgKCg41XlXrJhDdHQM9+8/ZMCPoxg4oDvr1m9Dq9USGBjM0WOnqFy5PGfPXmTRwmmYmpqwfcc+Ll3KeKzeD5l0LRVvzVctP8fByYEm9VuRkJDAyYv7MDc315nn9MlzfNOsIw0+q8Ps+ZP468/lhD8P5+jhk/T94Se9tuN/7yHXr9zgixb/u3FsZARzZixm9d8bdebt1qt99ndMvFEvX0ap/05ISNRp/rH4/+PFyMiIlas2MvKXyRmuL/megcoo9Udrj3mfpn6Db2japAHLl81i2vT56dY0PnQfRiiQewYGycYmNyFBISQkJFCjVhUKFsqfYp78BfISHBTKmhWbWbtqC2XLl+L82ctUqVYRtyIFAbCwtKBI0cLpbmv29IU6bf9HDp7gu3ZfqfcaXPO64OjkgM+pC3g2qou5uRlWuSxp0LDOm9thkW3+9x9SsWJZACpWKEORIoUAOHjIm6+/aoazsyMA9vZ2FErleErNMe9TtGrZHI1Gg5OTA7VrVcPH5yKFCuUnMDCYJUvXsGzZOipWKPt2duo98SbHQM5JUjMwQFs27mL52jnsPrCe61dvcMvvbop5PqlVhZ79OpMQn8DLl1H82GsEoSFhDOwzkjmLfsfc3AyAqRNmc+/O/TS35XfjDlcv+VKmfCkAjh46QbHi7uzYtxpIuvrs32M4ly5c5d9/DrP/6GYCAp5w+eI1XkS8eAt7L7Jiy5Y9tG/7LWd99nP27EX8biUdM76+t/h1zFT27lmLRmNEfHwC/fuP5MGDjId63LZtL59U+5jz5/5FURSGjZjAs2dBtG/fksGDehIfn8DLyJd06jLgbe+eQUv8QOoGMgay0JtVLkuiXkZjYWnB5l1/8/PAMVy97JvTxdKLjIEs0pLdMZAHuH2n97yz/Ndla1tvk9QMhN6mzhhDsRJFMTc3Y+O6He9NIBDibVI+kJqBBAOht77df87pIghhcAz9XoC+5AbyB8Ldw419RzapL9/7p+jasx2ff/kZB05s40HwZcpV+CiniylyWKPP6nHt6lFuXPfmp6F9Ml5AZEiLovfLkEnN4ANx97a/+qyCRqPh7LWD/LPrAJZWlvzQ4UemTB+dwyUUOU2j0TB71gQaN/2egIAnnDq5h5279uPreyvjhUWaDPsUrz8JBh+gWnWrc9//IY8CnuR0UYQBqVqlInfu+HPv3gMANmzYTvMvGkkwyKaEDyQcSDPRB6j5103YvnlPThdDGJh8+V15GPC/bLgBj56QL59rDpbow6Bk4j9DZpA1g8ePH7N48WLCw8OZNm0a9+/f5+zZs3zzzTdpLuPl5YWXl9c7LKVhMjU14bPG9Zj828ycLoowMEapPFH8AfYsf+fkBvJbtGDBAtq0aaNm1CxcuDAnTpxIdxlPT08mT57M5MkZP3b/IavvWZsrl30JDgrJeGbxn/Io4AkFC+RT/y6QPy9PnjzLwRJ9GD6UmoFBBoO4uDg8PDx0pmWUdlck+fKbptJEJFLlc/YiHh5FcHMriKmpKa1afcnOXftzuljvvQ8lHYVBnmFz587N06dP1WrtqVOnsLe3z+FSGT4LSwvq1PuEvTv/11zW+PMG+Fz1olKV8vy9bh6rNi1IZw3iQ5aYmMiAH39hz+41XL18mE2bdnL9ul9OF+u9l6goer8MmUGmo3j27BkLFy7k5s2b5MqVCxcXF/r374+zs7Ney0s6CvE6SUch0pLddBRtCn+l97xr7m/N1rbeJoO8gfzy5UtGjRpFTEwMiqJgaWnJ2bNn9Q4GQgjxrhj6vQB9GWQz0YIFC3jw4AEWFhZYWlpy/PhxtmzZktPFEkKIFOSewVs0aNAg5s6dS0BAAF5eXuzbt49ffvklp4slhBApfCjpKAwyGOTJk4cBAwYwbdo0Tp8+zS+//IKVlVVOF8tg2NjkZsHy6Rw+tYNDp3ZQqUr5FPN8UrMK+45s4sCJbWzauQwAB0d7tuxZgdfxrTRq+qk675JVs8njKk1w7zt98g7NmP4bN657c/7cv1SskHRvzcnJgSOHtnLxwgGaN//fqHdbNi8lb94876Ts77M32bU0ODiYsWPHMnDgQAYNGsSePUk9AyMjIxk3bhz9+/dn3LhxREZGqsts3bqVfv36MWDAAC5evJjl/TCoewaDBw/WeTAmMjISrVbLiBEjAPjjjz9yqmgGZeykYRw+cJwenQZhamqCpaXugOQ2NrmZ8McvtPu2B48fPcXRyQGAFt80ZdO67WzfspdVGxewb89BPBvV5eplX549DcqJXRFviD55h5o0/pRiHkUoWboW1apWYu6cSdSo9QXftW7BipUbWb9hO3t2rWbHjn00+7whFy5ckecQ9PAmewkZGxvTvn173N3diY6OZtiwYZQrV47Dhw9TtmxZWrRowbZt29i2bRvt2rUjICCAEydOMH36dMLCwhg3bhyzZs3KUld8gwoGw4YNy+kiGDzr3LmoVuNjBvYZCUB8fALx8bojjrX4til7d3rx+NFTAEKCQ9V5LSwsMDMzQ6vVYmxsTLee7enUpu+73QnxxumTd+iLLxqxcnXSWMWnz5zH1s4WV1cX4uMTsLS0wNz8f8dF/37d+PKrjjmyL++bN9n8Y29vr3ajt7S0JH/+/ISGhuLj48OYMWMAqFu3LmPGjKFdu3b4+PhQo0YNTE1NcXFxwdXVldu3b1O8ePFMb9ugmomcnZ3V18uXLzl37hznzp3j5cuX0pPo/xUqXIDQ4DCmzxnPP4c38vussVha6dYM3D3csLWzYeOOZew5uJ5vWjcHYNum3dT9tCarNi5g+pR5dOz6HZvW7yAmOiYndkW8QfrkHcqfz5WAh/+b51HAE/Lnc2Xtuq181rAeu3et5rdx0+nVsyMrV28iWo4LvbytG8iBgYHcu3cPDw8PwsPD1SBhb29PREQEAKGhoTg6OqrLODg4EBoamqX9MKiaQbI9e/Zw4MABqlatCsCff/6Jp6cnTZo0yeGS5TwTExPKlC/FqGETuXDuCmMnDaPPj135Y+Kc/81jbEy5CqVp3aIbFhbm7Ni3mvNnL3Hvzn06ftcbAFtbG3oP6MoPHQYwdeYYbO1sWDD3b877XMqpXRPZoE/eobTmiYh4QfMWHQCws7Nl6JDefNuqG/P/moq9vR0zZizg1Olzb6fgH4DMdC19PYeap6cnnp6eKeaLiYlh2rRpdOrUKd37pW/yMTGDDAYHDx5kwoQJWFhYAPDll1/yyy+/SDAAnjx+ypPHz7hw7goAu7fvp8+P3V6b5xmhoc+JjoomOiqa0yfPUbpMCe7dua/O8+NPPflz2kJafNOUyxevs23zbpau+pNWX3Z5p/sj3gx98g4FPHpCgYL/myd/gbw8fm2eUSMHMmnybL5r3YLz56+wdt1Wtm5ehudnLd/uDrzHMtNMlNbJ/1UJCQlMmzaN2rVrU61aNQBsbW0JCwvD3t6esLAwbGxsAHB0dCQk5H95yEJDQ3FwcMjCXhhYM1EyRVF0boBoNBrJrvj/ggJDePzoKe4ebkDS2AW3bt7RmWff3kNUrV4JY2NjLCwtqPBxWW773VXfL+JeiDyuLpw6cRZLS0sURYuiKJhbmL3LXRFvkD55h3bt2k/7tkkDIFWrWomI8AiePg1U3/fwKELefHk4euwUVlaWaLVJx4WFhfk73Zf3jaIoer/0Wdf8+fPJnz8/zZo1U6dXrlyZI0eOAHDkyBGqVKmiTj9x4gTx8fEEBgby5MmTFHnd9GWQNYP69eszcuRIdYd9fHz49NNPM1jqv2PUzxP5c8EUzMxMue//kMF9R9GuUysAVi3fwG2/uxw+eJx/vbeg1WpZu3IzN31vq8v/9Et/po6fDcC2zXtYsmo2XXq0Y9qkOaluTxi+V/MOGWs0LP97Pdev+9H9h/YALFy0kj17D9C48afc9D1OVHQ03boN0lnHuN9+ZtSvUwBYt34bWzYtpV+/rowZK7340pP4Bm8g37x5k6NHj1KoUCGGDh0KwPfff0+LFi2YMWMGBw8exMnJiUGDkr67ggUL8sknnzBo0CA0Gg1du3bNclJPg8tNpNVquXXrFqampty4cQOAUqVKUaRIEb3XIbmJxOskN5FIS3ZzE3kWbJTxTP/P6+G+bG3rbTK4moFGo2HFihVMmDABd3f3nC6OEEKky8Cup7PMIO8ZlC9fnlOnTn0wH7IQ4sP1oaSjMLiaAcCuXbuIjY1Fo9FgZmaGoigYGRnx999/53TRhBBCx4eStdQgg8GKFStyughCCKEXQx+0Rl8G2Uz022+/6TVNCCFymjQTvQVxcXHExcXx4sULnax8UVFRhIVJbxAhhOEx9JO8vgwqGHh5ebF7927CwsJ0ktZZWlrSqJH+3beEEOJd+VA6uhhUMChevDiffPIJp06dokmTJhw+fJjTp0/j7OxMrVq1crp4QgiRwodSMzCoewaLFi3C1NSUJk2acP36ddauXUvdunWxsrJiwYIFOV08IYRI4U0ObpOTDCoYaLVarK2tAThx4gQNGjSgevXqfPfddzx9+jSHSyeEECklKlq9X4bM4IJBYmIiAFevXqVMmTI67wkhhKF5k4nqcpJB3TOoWbMmY8aMIXfu3JiZmVGqVCkAnj59KmMgCyEM0odyz8CggsHXX39NmTJleP78OeXKlVMH49BqtXTu3DmHSyeEECkZ+r0AfRlUMABSHbszX758qcwphBA5T2vgzT/6MrhgIIQQ7xOpGQghhDD4XkL6kmAghBDZIM1EQgghpJlICCGE1AyEEEIgNQMhhBBAopKY00V4IyQYCCFENhh6mgl9STAQQohskHQUQgghpGYghBBCehMJIYRAehMJIYRA0lEIIYRA7hkIIYRA7hkIIYRAagZCCCGQ5wyEEEIgNQMhhBBIbyIhhBDIDWQhhBBIM5EQQgje/BPIFy9eZNmyZWi1Who0aECLFi3e6PrTonknWxFCiA+Uoih6vzKi1WpZsmQJI0aMYMaMGRw/fpyAgIB3sBcSDIQQIlu0iqL3KyO3b9/G1dWVPHnyYGJiQo0aNfDx8XkHe/GBNhMFhF7N6SIYDC8vLzw9PXO6GMLAyHHx5iTEPdJ7Xi8vL7y8vNS/PT09db6H0NBQHB0d1b8dHR25devWmyloBj7IYCD+R370IjVyXOSM10/+r0utKcnIyOhtFkklzURCCGEgHB0dCQkJUf8OCQnB3t7+nWxbgoEQQhiIokWL8uTJEwIDA0lISODEiRNUrlz5nWxbmok+cNIUIFIjx4VhMjY2pkuXLkyYMAGtVkv9+vUpWLDgO9m2kfKhPDEhhBAiy6SZSAghhAQDIYQQcs/AILRq1YpmzZrRoUMHAHbs2EFMTAytWrXSa/nnz58zf/58QkJCSEhIwMXFheHDhxMYGIifnx+1atXKVHk2bNiAhYUFzZs3z/S+iHdry5YteHt7o9FoMDIyonv37vj5+eHp6Ym5ubne6wkMDGTKlClMmzbtLZZWGDIJBgbA1NSU06dP06JFC2xsbDK9/IYNGyhXrhxNmzYF4P79+wAEBQXh7e2dajBITEzE2Ng4ewUXOcrPz49z584xZcoUTE1NiYiIICEhgT179lC7du1Ug4FWq0WjkQYBkZIEAwOg0Wjw9PRk9+7dfP/99zrvBQUF8ddffxEREYGNjQ29e/fGyclJZ56wsDDKlSun/l24cGEA1qxZQ0BAAEOHDqVu3bpYW1tz/vx54uLiiI2NZfDgwcybN4/AwEDMzc3p3r27umwyLy8vzpw5w5AhQzh16hR79+4lISGBYsWK0a1bNzmx5KCwsDBy586NqakpADY2NuzZs4fQ0FDGjh2LjY0No0ePpn379jRr1oxLly7RoUMHbt++zaFDhwD49NNP+fzzz3XW++zZM6ZNm0b37t2xtrZmyZIlREREYG5uTo8ePcifP/8731fxDigix7Vr1055+fKl0rt3b+Xly5fK9u3blfXr1yuKoiiTJk1SDh06pCiKohw4cECZMmVKiuUvXLigdOzYURkzZoyyefNmJSQkRFEURbl69aoyadIkdb5Dhw4pPXr0UF68eKEoiqIsWbJE2bBhg6IoinLlyhVlyJAhiqIoyvr165Xt27cre/fuVSZPnqzExcUpDx8+VCZNmqTEx8criqIoixYtUg4fPvx2PhChl+joaGXIkCFK//79lUWLFinXrl1TFEVRevfurYSHh6vztWzZUjl+/LiiKIpy584dZdCgQUp0dLQSHR2tDBw4ULl7967y7NkzZdCgQcqjR4+UoUOHKvfu3VMURVHGjh2rPH78WFEURfHz81PGjBnzbndSvDNSMzAQVlZW1KlThz179mBmZqZOv3XrFkOGDAGgTp06rF69OsWyFSpUYM6cOVy8eJELFy7w888/p9n2W65cOaytrQG4ceMGgwcPBqBMmTJERkYSFRUFwLFjx3BwcGDo0KGYmJhw9epV7t27x/DhwwGIi4vLUpOWeHMsLCyYMmUKvr6+XLt2jRkzZtC2bdsU82k0GqpXrw4kfedVq1bFwsICgKpVq+Lr60vlypWJiIhg6tSpDB48mIIFCxITE8PNmzeZPn26uq6EhIR3s3PinZNgYEA+//xzfv75Z+rVq5fpZa2tralVqxa1atVi8uTJXL9+ndy5c6eY79V2ZCWdR0wKFiyIv78/oaGhuLi4oCgKdevWpU2bNpkum3h7NBoNH330ER999BGFChXi8OHDKeYxNTVVm/PS+86trKxwdHTk5s2bFCxYEK1WS65cufj999/fVvGFAZEGXwNibW3NJ598wsGDB9VpxYsX58SJEwB4e3tTsmTJFMtdvXqV2NhYAKKjo3n27BlOTk5YWloSHR2d5vZKlSrFsWPHALh27Rq5c+fGysoKADc3N7p3786UKVMIDQ2lbNmynDp1ivDwcAAiIyMJCgp6MzsusuTx48c8efJE/dvf3x9nZ2csLCyIiYlJdZlSpUrh4+NDbGwsMTEx+Pj4UKpUKQBMTEwYOnQoR44cwdvbGysrK1xcXDh58iSQFEj8/f3f+n6JnCE1AwPTrFkz/vnnH/Xvzp0789dff7Fjxw71BvLr7t69y5IlSzA2NkZRFD799FM8PDxISEjA2NhY5wbyq1q1asW8efMYMmQI5ubm9OnTR+f9kiVL0r59eyZPnswvv/zCd999x/jx41EUBWNjY7p27Yqzs/Pb+SBEhmJiYli6dCkvX77E2NgYV1dXunfvzvHjx5k4cSL29vaMHj1aZxl3d3fq1avHiBEjgKQbyEWKFCEwMBBIanoaNmwY48ePx9zcnP79+7No0SK2bNlCQkICNWvWxM3N7V3vqngHJB2FEEIIaSYSQgghwUAIIQQSDIQQQiDBQAghBBIMhBBCIMFACCEEEgyEEEIgwUAIIQQSDIQQQiDBQAghBBIMhBBCIMFACCEEEgyEEEIgwUAIIQQSDIQQQiDBQAghBBIMhBBCIMFACCEEEgyEEEIgwUAIIQQSDIQQQiDBQAghBBIMhBBCIMFACCEEEgyEEEIgwUAIIQQSDIQQQiDBQAghBBIMhBBCIMFACCEEEgyEEEIgwUAIIQQSDIQQQiDBQAghBBIMhBBCIMFACCEEEgyEEEIgwUAIIQRgktMFEIbt3r17DBs2jOLFizNu3LicLo5BioyMZNmyZZw9exaAypUr06VLF3LlypXmMtHR0axfv54zZ84QHh5OkSJF6NSpEx4eHuo869at49SpU4SEhGBiYkKRIkVo3bo1JUqUeOv7JP57jBRFUXK6EMJwLV68GI1Gw5EjR5gwYQIFChTIsbIkJCRgYmJ41y8TJ04kODiYHj16YGRkxPz583FxcWHYsGFpLjNjxgwePHhAt27dcHR05OjRo+zevZsZM2bg4OAAwNGjR7G3t8fFxYW4uDh2797NyZMnmTVrFnZ2du9o78R/hTQTiTTFxcXh7e1NgwYNqF69OgcPHtR538/Pj7Fjx9K+fXs6duzIb7/9RmhoKACKorBz50769+9PmzZt6NmzJ2vWrAEgMDCQVq1acefOHZ31tWrVilOnTunM4+3tzdixY2nbti1eXl68ePGCmTNn0rNnT9q2bcugQYM4dOiQznrS2/bYsWNZsmSJzvxRUVG0a9eO06dPZ/ozCggI4OLFi3Tv3p0SJUpQvHhxunfvzvnz53n8+HGan+vp06dp06YNH330Ea6urrRq1QpXV1f279+vzlenTh3Kli1Lnjx5KFiwIB06dCA6Ohp/f/9Ml1OIjBjeZZYwGKdOncLZ2ZnChQtTp04dZsyYQZs2bTAxMcHf35+xY8dSp04dOnbsiImJCb6+vmi1WgDWrl3L/v376dChA6VLlyYiIoJ79+5lugxr166lffv29OrVC2NjY+Lj43F3d6dFixZYWlpy5coVFi5ciJOTE2XLls1w2w0aNGDJkiV06NABU1NTAI4fP46FhQUff/wxAAsXLuTYsWPplmvGjBk4OTnh5+eHhYWFTtNNiRIlMDc35+bNm+TLly/FsomJiWi1WnX7yczMzLhx40aq20tISMDLywtLS0vc3Nz0+/CEyAQJBiJNBw4coHbt2gCULl0ac3Nzzp49S/Xq1dmxYwdubm706NFDnT+5CSkmJobdu3fTsWNHPv30UwBcXV0pXrx4psvQuHFjqlevrjOtefPm6r/z5MnD1atXOX78OGXLls1w29WqVWPZsmWcOXOGmjVrAnDo0CHq1KmjNkG1bt1aZxupsbe3B+D58+fY2NhgZGSkvmdkZIStrS3Pnz9PdVlLS0uKFy/Oli1bKFSoEHZ2dnh7e+Pn54erq6vOvOfOnWPmzJnExcVhZ2fHqFGjpIlIvBUSDESqnj59ys2bNxkwYACQdIKrVasWBw8epHr16ty7d4+qVaumumxAQADx8fHqlXp2FC1aVOdvrVbLtm3bOHHiBKGhocTHx5OQkMBHH32k17ZNTU2pXbs2hw4dombNmgQEBHD79m169+6tzmNra4utra3eZXw1ECRTFCXV6cn69u3LX3/9Rc+ePdFoNBQpUoSaNWumqD199NFH/P7770RERHDgwAFmzJjB+PHj1WAkxJsiwUCk6sCBA2i1Wp2TZHJfg+Dg4HSXzahPgkajSTFfQkJCqvOam5vr/L1jxw527txJ586dKVSoEBYWFqxZs4aIiAi9tg1JTUVDhgwhODiYgwcPUrx4cZ0b45lpJrKzsyM8PFzn5K8oChEREekGFFdXV8aOHUtMTAzR0dHY29szY8YMXFxcdOazsLDA1dVVrd3079+fAwcO8O2332a4n0JkhgQDkUJiYiJHjhyhTZs2VKpUSee9OXPmcPjwYYoUKcLVq1dTXb5AgQKYmppy5coV8ubNm+J9GxsbAJ1mFH1vit64cYOPP/6YOnXqAEkn3idPnqjdODPaNkDBggUpVqwYXl5eHDt2jO+++07n/cw0ExUvXpyYmBj8/PzU+wZ+fn7Exsbq1QXUwsICCwsLIiMjuXTpEu3atUt3fkVR0gycQmSHBAORwvnz53nx4gUNGjQgd+7cOu/VqFGDf//9l8GDBzNq1CgWLFhAo0aNMDMzw9fXl/Lly+Pk5ESTJk1Yu3YtpqamlCpVisjISO7evctnn32GmZkZxYoVY/v27eTJk4eoqCi1t09G8uXLx4kTJ7hx4wa5c+dm7969BAYGUqRIESCpPT69bSdr0KABixYtwtjYmBo1auhsIzPNRAUKFKBChQosXLhQvX+ycOFCKlWqpHPz+Mcff6Rx48Y0btwYgIsXL6IoCvnz5+fp06esXLmSfPnyUa9ePSCph9OOHTv4+OOPsbe3JyIign/++YeQkBA++eQTvcomRGZIMBApHDx4kI8++ihFIAD45JNPWLNmDS9evGDUqFGsXbuWkSNHYmpqStGiRdWaRJs2bbC2tmbz5s2EhIRgZ2enXs0D9OrViwULFjB8+HDy5MlDt27dGD16dIZl+/rrrwkMDGTixImYmZlRr149ateuTUBAgDpPRtuGpKC2bNkyqlevjqWlZVY/KgD69+/P0qVLmTBhAgAff/wxXbt21Znn8ePHalMWJJ3s165dS0hICNbW1lSrVo3vv/9evYltbGzMw4cPOXToEC9evCB37twULVqUsWPHUrhw4WyVV4jUyENn4j8pNDSU3r17M2bMGEqWLJnTxREix0nNQPynJCQk8Pz5c9auXUuRIkUkEAjx/+QJZPGfcvPmTXr37s3t27d1npEQ4r9OmomEEEJIzUAIIYQBB4O5c+cyefJkvea9du0arVq10umt8TaNHTuWI0eOvJNtvSqtBG+ZMXz48CwlZBMZ27BhA4MHD9Zr3jfxXWbG3Llz2bRp0zvZ1rsgv4U3L8MbyHPnzlVPfMbGxjg6OlK1alVatWqFhYXFWytY586d9XqaFJISgy1cuDDVrpBv2vnz5wkODlZz9mRk7ty5vHjxIt10xu/SN998w4oVK6hSpYr6JHBW7Nu3jx07dvD8+XMKFChAp06dKFWqVLrLPHjwgCVLlnD79m2sra1p2LAh33zzjU7ahuvXr/P3338TEBCAvb09zZs313k+QB8bNmxQT3xGRkY4ODhQvnx52rRpoz7w9jY0b96cJk2a6DWvk5PTOztmHzx4wNmzZ5k7d+5b31Z65LfwPydOnGDr1q08efIEGxsbGjdurPOg4/Xr11mzZg2PHz8mNjYWZ2dnPv300wwfhswOvXoTlS1bln79+pGQkMCNGzeYP38+sbGx/PDDDynmTUxMRKPRpJuXRR9WVlZ6z2tiYvLOknft2bOHevXqZevgSc27ytVfqVIlFixYwMWLF1M8XayvEydOsHz5crp27UrJkiXZv38/EydOVFM0pCYqKopx48ZRqlQpJk2axOPHj5k3bx7m5uZ88cUXQNLV3qRJk6hfvz79+vXjxo0bLFmyBBsbmxTJ6jKSL18+xowZg1ar5d69e8yfP5/Q0FBGjBiRYt7kTKvZ/U6TnybWh0ajeWfH7N69e6lWrVqmflM56UP/LVy4cIHZs2fTuXNnKlSowKNHj1iwYAFmZmbqQ4kWFhY0adKEQoUKYW5uzo0bN1i0aBHm5uY0atQoy/ucHr0+cVNTU/XArVWrFlevXsXHx4cffviBDRs2cPr0ab744gs2b95MYGAgf//9N1qtlpUrV+Lj40NcXBxFihShQ4cOOonH/Pz8WLt2Lbdv30aj0VC0aFH69u2Lg4NDiquI69evs3r1ah48eIBGoyF//vz07NmTQoUKce3aNcaOHcvixYvVK7/Tp0+zYcMGnjx5gq2tLQ0bNuSrr75Sg1SfPn349NNPCQkJ4fjx41haWtK0adN0I29ERARXrlyhffv2OtP//fdfdu3aRXBwMBYWFri7uzNs2DA2b96s1qpatWoFwOjRo3F2dqZv375qnhk/Pz/at2/PZ599xpYtWzhw4ADh4eHkzZuX7777jipVqqRaHq1Wy9KlS7lw4QK//PILefPm5ezZs2zcuJGAgADs7OyoVasWLVu2VH9cGo2GihUr4u3tneUfwK5du6hbty6enp4AdOnShYsXL7J//37atGmT6jLe3t7ExcXRt29fzMzMKFSoEI8ePWLXrl00a9YMIyMj9u/fj729PV26dAGSnu69ffs2O3fuzHQwMDY2Vo9ZBwcHmjRpwvr164mLi+PEiRMsWbKEgQMHsnr1ah49esTUqVPJly8f69atw9vbm5cvX1KgQAFat25NhQoV1PU+evSIVatWcf36dbRaLYUKFaJHjx4UKlRI/S1MmzYNSLoiX758OXfu3EFRFPLkyUPHjh0pU6YMgYGB9O3bl0mTJqm/ievXr7Nq1Sru37+PlZUVNWvWpF27dup3N2bMGAoUKICVlRUHDhzAyMiIOnXq0K5duzQDmVar5eTJk/Tt21dnuj7Hf3BwMMuWLePKlSsAlCtXjs6dO+Po6Jjm5y6/hYx/C0ePHuXjjz9WT+p58uShRYsWbN++nUaNGmFkZIS7uzvu7u7qMi4uLpw5cwZfX9+cDQavMzMzIzExUf07MDAQb29vBg4ciImJCSYmJowdOxYrKyuGDRuGtbU1hw8f5rfffmPmzJnY29tnmA//VYmJifz+++/qFWNiYiL37t1L8wdw9+5dpk+fzjfffEOtWrW4c+cOCxcuVFMVJNu9ezetWrWiefPmXLhwgWXLllGyZMk0Uy3fuHEDExMTChYsqE67c+cOS5YsoU+fPpQsWZKXL1+qOXuaN2/Oo0ePiIyMpF+/fgBYW1urA8C8nqt/z5497Ny5kx9++AF3d3eOHTvGH3/8wZQpU1LksE9ISGDOnDk8ePCAcePG4eDgwMWLF/nzzz/VampwcDCLFi0iPj6eDh06qMt6eHiwdetW9W9fX18mTpyY6j4n++qrr/j6669JSEjg7t276tV8snLlynHz5s00l/fz86NkyZKYmZmp08qXL8/69esJCgrCxcWFW7duUa5cOZ3lypcvz5EjR7J9tWhmZoaiKOpxGx8fz5YtW/jhhx+wsbHB3t6eefPm8ezZM/r374+joyMXLlxgypQpTJo0CTc3N0JDQ/n1118pUaIEo0aNwsrKitu3b6d6zALMmjWLwoULM3HiRIyNjXnw4IHO/r8qNDSUSZMmUbt2bXr37s2zZ8+YP38+Go1G57s7duwYTZs2Zdy4cfj7+zN79mzc3d2pVatWquu9f/8+UVFROieWZOkd/4qi8Pvvv2Nqasqvv/6KkZERS5cu5ffff2fSpEmp1vzlt5Ako99CfHx8qmNZhISEqL+F1927d4+bN2/SsmXLdMuWHZn+dd2+fZvjx49TpkwZdVpCQgJ9+/ZVr8SuXr2Kv78/S5YsUQ/+7777jnPnznH06FG+/PLLdPPhvy46OpqXL19SuXJlNd97/vz50yzjrl27KF26tHoFki9fPp48ecL27dt1gkG5cuXUalmTJk3Yu3cvV65cSTMYBAUFYWtrqxOEgoODMTc3p3LlylhaWuLs7KwerMbGxpiZmenUrF71eq7+nTt38sUXX6g/7NatW+Pr68uOHTvo37+/Ol9sbCxTpkwhKiqK3377DWtrawC2bt3KF198Qf369YGkzJht27blzz//pH379uoP2MHBgdDQUBITEzE2NqZo0aL8/vvvaX6egLqNiIgItFptitw9dnZ26hVkap4/f64O5/jqMsnvubi48Pz58xSpp21tbUlMTOTFixdZTtv86NEj9u/fj4eHh5p6QqvV0qVLF/Uk+fTpU44fP87cuXPV6n3jxo25fPkyXl5edOvWjX379mFubs6gQYPUwJTa4DXJgoOD+eKLL9Rj9fWxCl61b98+7Ozs6NatGxqNhgIFCtC2bVsWLlxI69at1eytybWV5G0fOHCAq1evphkMgoKCMDIySvX4S+/4v3LlCv7+/vz555/qyal///7079+fK1eupAjayfsrv4WMfwsVKlRg2bJlXLp0ibJly/L06VN27doF/O+3kKxnz55ERESQmJhIy5YtM33/LDP0CgYXL16kffv2aLVaEhISqFKlilqVh6QP9NUv+O7du8TFxaXIzxIfH8+zZ88A0s2H/zpra2vq1avHhAkTKFOmDGXLlqV69epptsk9evSIihUr6kwrWbIkmzZtIioqSm07fT3Hi729PeHh4WmWIy4uLkVEL1eunFrVLV++POXKlaNatWp65bt5tcksKiqKsLCwFJkuS5YsyYULF3Sm/fnnn9jZ2TF69GidNuq7d+9y+/Zttm/frk5TFIW4uDieP3+unkyTr5Lj4+PVH2l6J6rUvH5lmFH+/rSWyer2MhIQEKBzzJYuXVrnwsPY2FjnCvPevXsoisLAgQN11pOQkKBe+Pj7+1OyZEm9ayiff/45CxYs4MiRI5QtW5Zq1aqleRHz6NEjihcvrnOhUbJkSRISEnj69Kl6rGblmDU2Nk61Fp3eugICAnBwcNA5MeXJkwd7e3sCAgJSDQbyW/jfdtI7Xhs0aMDTp0+ZOnUqiYmJahPdxo0bU3xPv/32m5oVd/Xq1bi4uKTIs/Wm6HVUlypVih49emBsbIy9vX2KH8PrN82So+Vvv/2WYl1ZTQrWu3dvmjZtysWLFzl79ixr165l6NChOu25ydL7Ml6dbmxsnOK99E5QuXPn5uXLlzrTLC0tmTJlCr6+vly+fJlt27axdu1aJk2alOJK+HWv5+p/vXxpqVixIkePHuXmzZuUL19ena7Vavn2229TzWr5ai+ayMhITE1N1e8tM1VjGxsbNBpNilG8wsPD0830aWdnl+oyye+lN4+xsbF6NaYvV1dXhg8fjkajwd7ePkUQNzEx0fnhJR8zkyZNSnF8J9duM/t8ZqtWrahduzYXLlzg0qVLbNy4kR9++EEdge1Vb+uYtbGxISEhgdjY2BTHW2bXlVp5XiW/hSQZ/RaMjIxo164dbdq0UUfKS65JODs768ybHIwLFSpEeHg4GzduzNlgYG5unqlo6e7uTnh4OEZGRuTJkyfVedLLh58WNzc33NzcaNGiBRMnTuTIkSOpBoMCBQqkGEv2xo0bODo6ZitDZZEiRYiIiCAiIkLngDI2NqZMmTKUKVOGVq1a0a1bN86fP4+npycmJiZptim/ysrKCnt7e27cuKHTBHfjxo0UzWcNGjTA3d2d33//naFDh6o/And3dx49epThd/XgwQOdNuTMVI1NTExwd3fn8uXLOj+0K1euUK1atTSXL168OKtXryYuLk49uV6+fBl7e3v1B1CsWDF8fHx0lrt8+TLu7u6Zvl9gYmKSqWPWzc0NRVF4/vy5zuf/qiJFinDs2LFM3b/ImzcvefPmpWnTpixatIiDBw+mGgwKFCjAyZMn0Wq1apBKvkeV1m9I3/2CpCv910eNS0+BAgUIDQ0lMDBQPSE9e/aMsLCwNJtzQX4LkPFvIZlGo1GD5PHjxylevHi6QSS5BvO2vJWHzsqWLUuJEiWYOnUqFy5cIDAwED8/PzZs2ICvry+QdEPJ39+fBQsW4O/vz+PHjzlw4ECqo2gFBgayevVqbt68SVBQEFevXuX+/ftpHpTNmjXj+vXrbNiwgcePH3Ps2DF27dqV7T66RYoUwdbWVifQnDt3jj179nDv3j2CgoLw9vYmOjpabQ5wdnbm4cOHagrj9AYmad68OTt37sTb25vHjx+zfv16fH19U9ygAvD09KRjx478/vvvXL58GUjqN338+HHWr1/PgwcPePToEadOnWLVqlU6y964cUPnKiq5apze69Ur82bNmnH48GEOHDhAQEAAy5YtIzQ0lIYNG6rzrFmzRqdmWKtWLczMzJg3bx4PHjzg9OnTbN++Xe1JBPDZZ58RGhrK8uXLCQgI4MCBAxw+fDjV/X/T8uXLR61atZg3bx6nTp3i2bNn3Llzhx07dqgPJjVq1IiYmBimT5/O7du3efr0Kd7e3qkOzBMXF8fixYu5du0agYGB3Lp1K9WTWbJGjRoRFhbG4sWLCQgI4Pz586xevZrGjRunetWsLxsbG4oUKZLi4igjZcuWxc3NjT///JO7d+9y584dZs+eTZEiRdIMlvJb0O+3EBERwf79+wkICMDf359ly5Zx8uRJOnXqpM6zd+9ezp07x5MnT3jy5AkHDx5k586dej/flBVvpTOvkZERw4cPZ926dSxYsIDw8HDs7OwoUaKEWsVxc3NLNx/+q8zMzHjy5AnTp0/nxYsX2NraUrt2bb788stUt+/u7s6gQYPYsGEDW7duxc7OjhYtWqg3y7JKo9FQv359vL291fsduXLlwsfHh02bNhEbG4urqys9e/ZUHzrx9PTk+vXrDBs2jJiYGLU7XWqaNGlCdHQ0q1ev5vnz5+TLl4/Bgwen6D2RrGHDhmqvj+Qms+RufDt37sTY2Ji8efOqA6ZAUq+Vmzdvqj06sqJGjRq8ePGCLVu2EBYWRsGCBRk+fLjOfoWFhan3hyDpam/UqFEsWbKE4cOHkytXLpo1a0azZs3UeVxcXBg+fDh///232s20c+fOOjcWk7sRjx49Wh33+E3p3bs3W7ZsYdWqVeo4Ax4eHurJz8HBgbFjx7Jq1SrGjh2LkZERhQoVonv37inWpdFoePnyJXPnzuX58+fkzp2bSpUqpeiWnMzBwYHhw4ezatUqfvrpJ3LlykXNmjX5/vvvs71fnp6eHDx4kM8//1zvZYyMjBg6dChLly5lzJgxQFKA6NKlS5rNN/Jb0O+3AHDkyBFWrlwJJNWax4wZg4eHh/q+Vqtl9erVBAUFodFocHV1pU2bNjpB5k2TRHWZFB4ezqBBg5g4cWK2qu85ZeXKlURFRb23GTsPHTrEmjVrmDlzpjrUpUhffHw8P/74I3379s3wydj/kvf9t/CmGWxuIkNla2tLr169CAkJyemiZImtra3aNfF9dOHCBdq2bSuBIBNMTU3p06cPkZGROV0Ug/K+/xbeNKkZCCGEkJqBEEIICQZCCCH4jweD18dM0HcMhQ0bNjBv3ry3WbQ09enThx07dmR5+RUrVrB06dI3WCKRltfH2dB33I0nT57www8/EBUV9S6K+U5kZnyS1Jw7d46hQ4fq9ZyCyJq3nyc2Fa+OkZD8hGilSpX4/vvvM/2k6bsWHh7O7t27mTp1ql7zHz58mCVLlqjdyHJaixYt6NevH59//nm2ekNlZdyByMhIli1bxtmzZwGoXLkyXbp00bkZHBwcrPbPNzMzo2bNmnTo0CFTD50ldz9NZm1tTeHChWndujUlS5bM5J6+e2vXruWzzz7L0ZTTqWVVzUkff/wxGzZswNvbO1tP4C5btoybN2/y8OFD7Ozs9BrjQVEUNm7cyIEDB4iMjKRYsWJ07dpVJ2FlfHw8K1eu5Pjx48TFxVGmTBm6deuWboZXQ5NjNYOyZcuycOFC5s6dS8+ePTl37hyLFy/OqeLo7cCBA3h4eLzxbqXpPYDzJtnY2FCuXDn279+f5XUkjztQokQJpkyZQosWLVi2bBmnTp1Kd7nZs2dz7949RowYwciRI7l37x5//vmn+r5Wq2XSpEnExMQwduxYBgwYwOnTp1mxYkWWyjl9+nQWLlzImDFjsLGxYfLkyenm8TEEwcHB+Pj4qAnWDN27Om4B6tevz969e7O1DkVRqFu3bqYCyvbt29m1axedO3dm0qRJ2NjYMH78eKKjo9V5li9fzunTpxkwYABjx44lOjqayZMnv1c1mRypGYDuGAmOjo7UqFGDw4cP68xz6NAhduzYQWBgIE5OTjRs2JCmTZuqj+tHRUWxevVqfHx8ePnyJS4uLrRs2VJ9EGTJkiXcuHGDFy9ekCdPHp0shll1/PhxGjRooDMtrbEWXrx4oTYnJWdQ/fbbb2nVqhV9+vShbt26BAcHc+bMGcqVK8egQYMyHIfhdUePHmXJkiX069ePypUrExAQwMqVK/H19cXMzIwyZcrQqVMnnUSClStXVlMGZ0VWxh0ICAjg4sWL/Pbbb2oCsu7du/Prr7/y+PFj8uXLx6VLlwgICNDJHNq2bVsWLFjAd999l+krZRsbG2xsbLCzs+Obb77h5MmT3Lp1i8qVK6tlyuizOnz4MDt37uTJkyfkypWLChUq0KdPHyApO+7hw4d59uwZVlZWVKxYkfbt22er2+uJEycoVKiQThLG5NrlTz/9xPLlywkMDMTDw4NevXrpJJL7999/2bFjB8HBwTg5OfHll1+qefZTExwczNKlS/H19SU+Ph4nJydatmxJzZo11fEPhg8fDkDp0qUZM2aMOs5IyZIl+eeff0hISGDx4sU8ePCAv//+mxs3bmBmZkblypXp3Llzmt+Zv78/EydOpH79+nz//fdERUVlOP5J5cqVWbp0KU+fPs10Mrlkycfsjh071KeV06MoCnv27KFFixbqsd23b1+6deuGt7c3DRs2JCoqioMHD9K7d281gV/fvn3p06cPly9fTjVljiHKsWDwqmfPnnHx4kWdxFleXl5s2LBBTTP84MEDFixYgImJCY0bN0ZRFCZNmkRkZCS9e/cmb968PH78WM3dER8fj7u7Oy1atMDS0pIrV66wcOFCnJycUqRJ1ldkZGSKHC/pjbVQokQJOnXqxNq1a9Ur4FeT+u3evZuvv/6ayZMnoyiK3uMwJNuzZw8bN27k559/pnTp0oSFhTF69Gjq169P+/btSUxMZO3atUyZMoUJEyaoQdTDw4PQ0FCdH1VGgaFUqVLqKGFZGXfAz88PCwsLnUyUJUqUwNzcnJs3b5IvXz78/PzInz+/zomwQoUKxMfHc/fu3TTTIGQkNjaWQ4cOAahl0+ez+vfff1m+fDnff/89lSpVIiYmRieflpGREZ06dcLFxUU9sS5dujRbT7TeuHEj1bEHEhIS2LZtG7169cLU1JS5c+eyaNEiRo4cCcCZM2dYunQpHTt2pFy5cly6dIklS5ZgZ2enBr/XLV68mPj4eEaPHo2VlRWPHz9W35s4cSIjRoxgxIgRuLm56Xyn169fx8rKSj0eYmNjmTBhAkWLFlV/kwsWLGDevHkMGTIkxXZ9fX2ZOnUq33zzDc2aNVN/y+mNfwJJQ4Xa2tpy/fp19bidOHGimuImLdlpog0MDOT58+c6x7uZmRmlSpXi5s2bNGzYkLt375KYmKiT1sLJyYn8+fPj5+cnwSAjr6bFTj6BvzroxObNm2nXrp0ajV1cXHj27Bn79u2jcePGXLlyBT8/P6ZNm6bme3m16cbBwUEnF1GePHm4evUqx48fz3IwCA4ORlEUnbz6GY21kHxllFoO91KlSumk1Jg9e7Ze4zAArF+/Hi8vL3799VeKFCkCJF2xFy5cmHbt2qnz9e3bly5dunD37l31cffk8gcFBallzig516uDsmRl3IHk7Iyv1nCMjIywtbVVsz4+f/48RaKu3Llzp5oZUh/JV/BxcXEoikLRokXVgKLPZ7V582aaNm2qkzLj1RP1q+kdXFxcaNeuHVOnTqVPnz5ZHkIzKCgo1ZQLiYmJdO3aVR0/4YsvvmDevHlqYrvkvDXJKVfy5cvH3bt32b59e5rBIDg4mGrVqqnbe7WWkZyIMXfu3CmOXVNTUzUoQdKFW0xMDP369VMTQXbv3p2xY8emuIo/d+4cs2fPpkuXLtStWxdIuseT0fgnyRwcHAgMDFT/7tmzJ3Fxcel/qNmQfNy9/hnY2toSFhamzqPRaFKMZ/3qsf0+yLFgkJwWOy4uDi8vL549e0bTpk2BpEROISEhLFy4kEWLFqnLaLVaNcXuvXv3sLOzSzPxl1arZdu2bZw4cYLQ0FDi4+NJSEjIVj6b5IPu1XTImR1r4VWv35jTdxyGvXv3Eh0dzaRJk8ibN6867927d/H19U31Kv/p06dqMEj+wb36I8pqtft16aUdTu2911M365PGWV/JV7z+/v6sWbOGPn36qFe4GX1Wzs7OhIaGpnvhcPXqVbZu3cqjR4+IiopSx05IbSAffaU2ZgYkHXOvDqRjb29PYmIiUVFRWFtbExAQkKIJtGTJkurN+tQkZ1K9ePEiZcuWpWrVqqnWSl5XqFAhnTI+evSIwoUL62QELlGiBEZGRgQEBKjH1t27d/njjz/o37+/TpZPfcY/SWZmZqZz3Gb1cxYp5VgweDUtdpcuXRg7diybNm2iVatW6k2XH374IcUAF8kyenB6x44d7Ny5k86dO1OoUCEsLCxYs2ZNht360pMc+V++fKlz9ZuZsRZe9Xo2Sn1z2pcoUYJLly5x/Phxvv32W53lK1asqFPDSvbqFXdyWoJX03BnppkoK+MO2NnZER4errOPiqIQERGhls3Ozi7FcIEvXrxIdTQpfbi4uGBjY0O+fPmIi4vjjz/+UIdyzOizio2NTXfdQUFBTJo0iQYNGtC6dWusra25d+8es2bNytZNVRsbmxRjZgApahrJn2FGNyjTC6Kffvop5cuX58KFC1y+fJlffvmFFi1aqDXTtGTmuH2Vi4sLtra2HDp0iMqVK6sBJTPjn0RGRuoct2+7mejV0fhevcB7dcwCOzs7tFotL1680ClbeHj4e5ULyiDuGUDSjdWJEyfi6emJg4MDDg4OPHv2TK1Kvs7d3Z3nz58TEBCQau3gxo0bfPzxx2qvAUVR1JuAWZUnTx4sLS1T3WZaYy3om8Md9B+Hwd3dnWbNmjF+/HgANSAUKVKEkydP4uTklG5XzIcPH2JsbEyhQoXUaZlpJsrKuAPFixdXR2xKDvB+fn7ExsaqfxcvXpwtW7YQEhKidsm7fPkypqamel2xpqdOnTps3ryZffv20axZsww/K0tLSxwcHNIc4vHOnTskJCTQqVMn9UR9/vz5bJURko6jgICATC+XfOy8OlZCeimzkzk6OuLp6Ymnpyfbtm1j7969tGrVSv1M9Dl2CxQowKFDh4iOjlaP05s3b6Iois72ra2t+emnnxg3bhx//PEHQ4YMUb/bjMY/gaRa09OnT3WOhbfdTOTi4oKdnR2XL19Wa9ZxcXHcuHFDbWJ0d3fH2NiYy5cvq8N0hoSEqKPXvS8M5qGzjz76iIIFC7JlyxYAWrZsqXbpevz4MQ8ePODIkSPq4NVlypTBw8ODadOmcfHiRQIDA7l8+TJnzpwBktpMr169yo0bN3j06BFLlizRaWvMCo1GQ9myZXVO2BmNteDs7Ex8fDyXL18mIiIi3SvOzIzD4OHhwS+//MKuXbvYvHkzkJQTPyoqipkzZ3Lr1i2ePXvG5cuXWbBggU43OF9fX0qVKqVzhZdRDvdXq+P6jDtw5swZfvzxR3XA8wIFClChQgUWLlyIn58ffn5+LFy4kEqVKqnNH+XLl6dAgQLMmTOHe/fucfnyZVauXEmDBg2y3edeo9HQtGlTtm3bRkxMjF6f1ddff82ePXvUY9Df35+dO3cCSYPWKIrC7t27CQwMxNvbm927d2erjMmfwa1bt0hMTMzUcl988QXHjh3jn3/+4cmTJ+zduxdvb+90x/BYtmwZFy9e5NmzZ/j7+3Pp0iX1uLW1tcXMzIxLly7x/PnzdB+Aq127NhYWFuqg9NevX2fhwoVUrVo1RfOjjY0No0aNIiQkhD/++IP4+Hi9xj+BpI4LpqamOq0FDg4OGR67r3r69Cn+/v6EhYWRkJCAv78//v7+am0uNDSUH3/8UT2PGBkZ0bRpU7Zv387p06d58OAB8+bNw8LCQj3xW1lZ8emnn7Jy5UouX77MvXv3mDNnDoUKFUr1QsJQGUzNAJJOhvPmzePLL7+kQYMGmJubs3PnTtauXYuZmRkFChRQb5BpNBpGjBjBypUr+fPPP4mJiVG7lkLSDzkwMJCJEydiZmZGvXr1qF27dpauul7l6enJvHnz6NChAxqNJsOxFkqUKEHDhg2ZNWsWL168ULuWpiaz4zAkB4TkGsI333zDuHHjWLNmDRMnTiQuLg4nJyfKly+v08Z7/PjxDJsC0qPPuANRUVE8fvxYp8mkf//+LF26lAkTJgBJDxK92k6s0WgYPnw4ixcvZtSoUZiZmVGrVq0UTVitWrVK93NMS/369dmwYQN79uzh66+/zvCz+uyzzzAxMWHnzp2sXr0aa2tr9Z5O4cKF6dSpE9u3b2fdunWUKFGC9u3bM3PmzEyV6XWVKlVST8Kpje2RlqpVq9K5c2d27tzJ33//jZOTE127dk3z5jEk1ZaXLl1KSEgIFhYWlC1bVm02MzY2pnPnzmzatImNGzdSqlQpdVyD15mbmzNy5EiWL1/O8OHDdbqWpsbGxoZff/2V3377jWnTpjF48OAMxz+BpOO2Vq1a2RrsZ/78+Vy/fl39+6effgJgzpw5uLi4kJCQwOPHj3WC35dffklcXBxLlizh5cuXeHh4MHLkSJ3aeseOHTE2NmbmzJnqQ2fZ6UiQEyRraRaMHDmSRo0avbWxSN+m8+fPs3LlSv74448UY+C+DwIDA+nXrx9jx459L54mzor9+/dz6tQpfv3115wuisGIiIjgxx9/ZPLkyTq9nsSb8/6ELQPSvXv39+rJwlfFxMTQu3fv9zIQQFIwq1OnzgcbCCCp9lmmTJkPKjdRdgUGBtKtWzcJBG+R1AyEEEJIzUAIIYQEAyGEEEgwEEIIgQSDLImMjOSHH37g6dOn73zbhw8fznK2UUh6KrJr166EhIS8wVIJQ5CQkMCAAQN0uk6+7+R4f3cy9ZzBq4PSvGrq1Km4ublx/fp1du7cyd27dwkLC6N3797Uq1cvw/UeOHCAffv28eTJEzQaDc7OzlSuXJnvvvsuM8V7Z7Zu3UrFihX1zufTp08fGjVqlO4DQO+Kra0tdevWZcOGDfTq1Suni/NGbNiwgU2bNqWYPmTIEKpWrcrDhw/ZsGED9+7dIzAwUO9nFM6cOcP27dt59OgRiYmJODo6UrJkSXr27Pk2diPbvLy8cHBwoHTp0jlajg/xeM/K4DXJmWaPHDlCaGgo+fLlo23btjppavr06UNQUFCKZStWrKimD39XMv3QWdmyZVOk6E3O2RMTE0PBggWpW7cuc+bM0Wt9Bw8eZNmyZXTs2JGyZcuSkJDAw4cP8fPzy2zR9JbcLTQrD4TExsZy8OBBfv75Z4MpU2bVq1eP4cOH0759e4MfWU5f+fLlS/FQVHLqkdjYWJydnalWrRrr1q3Ta31XrlxhxowZtGzZkt69e6PRaHj06FGKNBxvWlopwPXxzz//8PXXX7/hEr0d79vxvnz5cs6ePcuAAQOwtrZmxYoVTJ48mSlTpqS5D+vWrePYsWP06NGD/Pnzc+nSJX7//XfGjx+vZhqeNGmSTjf1sLAwhg0bppPI713J9FH36qA0r6tUqZL61KQ+w8kBnD17lqpVq9KwYUN1WoECBVJ8GOfPn2fTpk3cv38fc3NzihcvzqBBgzAzMyMyMpLly5dz7tw54uLiKFmyJJ06dVKHpUseHGTgwIGsXr2aR48eMXXqVPLly8e6devw9vbm5cuXFChQgNatW6ebYO7ChQsYGRnpPBKfkJDAihUrOH36tPoUcq1atWjbti1jxowhKCiIVatWsWrVKiDpSjatMjk4OKS7L6+LjIxkypQpWFhYMHjwYMzNzdmxYwdeXl6Ehobi6urKl19+qfOAXKFChbC3t+f06dMpBurRV6tWrejevTuXL1/mwoUL2Nra0qpVK53trF69mjNnzhAcHIydnR2ffPIJrVq1UvMcbdiwgdOnT/P111+zbt06wsPDKVOmDD179tRJ+KUPY2PjNI9LDw8PNa9McjqTjJw7dw4PDw+dk2vevHlTPNHr5+fH2rVruX37NhqNhqJFi9K3b18cHByIj49n9erVHD9+nKioKNzc3Gjfvr36jETy8JzDhg1j48aN+Pv7M2TIECpVqpThd/i6O3fu8OTJEz7++GN1WvLQlYMGDeLff//l5s2bODs707lzZ500CdevX2fVqlXcv38fKysratasSbt27dIMSv+14z2rg9ccO3aML7/8Uj0nfvbZZ1y+fJmdO3fSv39/gBTH+cGDB7G0tHw/gsGbZmdnx7Vr13j27FmaSaouXrzI1KlTadGiBb179yYxMZFLly6pmUvnzZvH48ePGTp0KNbW1qxdu5aJEycya9Ys9cQTHx/Pli1b+OGHH7CxscHe3p558+bx7Nkz+vfvj6OjIxcuXGDKlClMmjQp1ZzykJTXx93dXSdL4969e/Hx8WHAgAG4uLgQEhKiDhQyZMgQhg4dSv369VOMEZxamf78888M9yVZWFgY48ePJ3/+/PTv3x8TExPWrl3LqVOn1Nz3fn5+LFiwAGtra530Bh4eHly/fl39cWzZsiXDE+WIESN0sjBu2rSJNm3a0KZNGw4ePMhff/1FqVKlcHZ2BpLSFPTq1QsHBwcCAgJYtGgRJiYmOs1/gYGBnDhxgiFDhhAbG8vMmTNZt24d3bt3T7csb5udnZ2ajyitY8Hf35+xY8dSp04dOnbsiImJCb6+vuqV3qpVqzh58qQ6ItmuXbuYMGECs2fP1sl6u3r1ajp06ICrqyuWlpasW7dOr+/wVb6+vri6uqaaiHHdunW0a9eObt26sXnzZmbOnKnm1wkNDWXSpEnUrl2b3r178+zZM+bPn49Go0k1oyv89473rA5eEx8fnyIduZmZWYrMvMkUReHgwYPUrl07Wyk3sirTwSB5UJpkr6Y2zoqWLVty//59+vXrh6urKx4eHpQvX56aNWuqVyabN2+mevXqOieRwoULA/DkyRPOnj3LmDFj1LbSfv360atXL44dO6Z++VqtVh01DZISVh0/flxniMXGjRtz+fJlvLy86NatW6rlDQoKSnEFGhQURN68eSlVqhRGRkY4OTmpNQdra2s0Gg0WFhYplnu9TPruS3L5x48fT/ny5enatSsajYaYmBh27drFL7/8op60XVxcuH37Nvv27dP5cdjb23Pnzh31788++4waNWqk+129nju+Tp066hVY69at2bNnD76+vmoweDW9touLC1999RU7d+7U+R61Wi19+vRRE9F5enqqo5JlRkBAgM5x6ezszPTp0zO9nmSNGzfG19eXn376CUdHRzw8PChbtix16tRRR6vbsWMHbm5u9OjRQ10uOdFbTEwM+/fvp2fPnurn3r17d65du8a+fft0PoOWLVuqJ5rMfIevSq59pebzzz9XazRt2rTh6NGj+Pv7U7JkSfbt24ednR3dunVDo9FQoEAB2rZty8KFC2ndunWqJ6X/2vGe1cFrypcvz549e/joo49wdXXl6tWrnDlzJs3sBZcvXyYwMDDLtfXsynQwSB6UJtnr0Tuz7O3tmTBhgprtMDmb5e7duxk3bhzm5ubcu3cvzVTWjx49wsjISCdVrJWVFYUKFdJJSmdsbKxzhXfv3j0URWHgwIE660tISEh3eMX4+PgU+1yvXj3Gjx/PgAEDKFeuHJUqVaJChQoZtoe+XiZ99yUhIYFRo0ZRrVo1naAVEBBAfHw8EydO1NlOYmKieoJO9vogIdbW1pluT301BbaxsTE2NjY640WcOnWK3bt38/TpU2JiYtBqtSl+CE5OTjoZSe3t7bM05oSrq6vODbfsptuwsLBg+PDhPH36lGvXrnHr1i3Wrl3Ltm3bmDhxInZ2dty7d4+qVaumuvyzZ89ITEzUaU7UaDQUK1YsRbLEVwc5ysx3+Kq4uLg0f4vJF07wv1HuwsPDAdQ0y68eqyVLliQhIYGnT5/qLJvsv3q8Z1bnzp2ZP38+AwcOVNNz16tXL8VY78kOHDhA0aJF06yJvm2ZDgavDkrzJhUqVIhChQrRuHFjbty4wa+//srJkycz7I2UXjaNV5tyTExMdA7W5AE5Jk2alKJtNL0Alzt3bnVwmGTu7u7MnTuXixcvcvXqVebOnUvhwoX55Zdf0v2BpFYmffcleVCSoKAg9cBPXv7nn39OMdLa6yfH1wcJyUoz0eufm5GRkXqy9/PzY+bMmXz77bd07NiRXLlycfbs2RQDjaS3jswwMTF5K8dlchrkBg0a8PXXXzNgwAD279+fYW+k5O9Cn0FfXr36zsx3+KrcuXNz7969VN97dblXBxZK/n9mR5f7rx3vWR28xsbGhp9++om4uDgiIyOxt7dn9erVqeZXCg8Px8fHJ8Vob+9Sjt8zSM2rVW1IGrTl6tWreHp6pjqvoij4+fmpVc2oqCgePHiQYhjAV7m5uaEoCs+fP8/UQOtubm6pdq9NvunzySefUK9ePUaOHMnTp0/Jly+f3gPcZGZfevfuzdy5cxk7dixjxozBycmJAgUKYGpqSlBQUIb79PDhQ50DOSvNROm5efMmDg4OOk1FqXWhe584OztjZmaW4rhMjaurKyYmJty4cUO9F6bVarl16xY1a9ZMcxuZ+Q5fVaRIEf755x91TGR9FShQgJMnT+osd+PGDUxMTNIdaOa/dLxnd/AaMzMzHBwcSEhI4PTp06neHD506BCmpqbpHhtv2xsNBjExMeqDWIqiEBwcjL+/P9bW1mmOCbxo0SLs7e0pU6YMjo6OhIWFsWXLFszNzdU791999RVTpkzB1dWVWrVqoSgKly5domHDhmoPj0WLFtG9e3dy5crF2rVrsbKyUr+41OTLl49atWqpYxMUKVKEyMhIrl27Rp48eahWrVqqy1WoUIHVq1fz4sULtQ1x165d2NnZ4ebmhomJCd7e3lhaWqp9kJ2dnblx4wahoaGYmJik2VMmM/ui0Wjo06cPc+bMYcyYMeoP5IsvvmDlypUoikLp0qXV0cU0Go0aTGNjY7l79y7ff/+9ur43XW3OmzcvoaGhHDt2jOLFi6vDdOaEhIQEtdkhLi6O58+f4+/vj4WFRZq1iQ0bNhAXF0fFihVxdnbm5cuX7N27l5iYGLX9vXnz5owcOZIFCxbQqFEjzMzM8PX1pXz58jg5OfHZZ5+xZs0acufOjYuLC7t37+b58+c0atQozbJaWlrq9R2+7qOPPiI+Pp4HDx5kqpmhUaNG7Nmzh8WLF9O0aVN1sKbGjRuneRPzv3a8vzp4jY2NDblz52bFihUpBq/57bff8PDwoE2bNkDSYDyhoaG4ubkRGhrKxo0bURRFHeskWfKN4xo1aqQY5vNdeqPB4M6dO4wdO1b9e8OGDWzYsIG6devSp0+fVJcpV64chw4d4t9//+XFixdYW1vj7u7OL7/8oo6AValSJYYOHcrGjRvZsWMHlpaWFC9eXO2t0Lt3b5YvX87UqVPV7mkjRozI8H5G79692bJlC6tWrSIkJARra2s8PDzSvcooVKgQHh4eHD9+XB10xsLCgp07d/LkyROMjIxwc3NjxIgR6o+pVatWLFq0iH79+hEfH8+GDRvSLZO++6LRaOjbty9z5sxh7NixjB49mtatW2Nra8vOnTtZvHgxlpaWuLm56RyAPj4+ODk5vdXxWStXrkzz5s1Zvnw5cXFxlC9fntatW7N48eJMrSe5++Xo0aP56KOPslSW0NBQdRATSGrP9/LyonTp0mkO2FK6dGn279/P3LlzCQ8Px9LSkoIFC/Lzzz+rV7Fubm6MGjWKtWvXMnLkSExNTSlatKh647Jt27YA/PXXX7x8+ZIiRYowcuRInZ5EqdHnO3xd7ty5qVq1KseOHctUMHBwcGD48OGsWrWKn376iVy5clGzZk2dE+fr/ovHuz6D1zx79kznIbT4+HjWrVtHYGAgFhYWVKxYkb59+6bo8XXt2jWePn2qdjfNKZLCOgsuXrzIsmXLmDFjxns1klGy4cOH8/nnn6dbczIUhw4dYs2aNcycOTNb41f/Fzx8+JCxY8cye/bsbA8T+iF5n473nPT+nckMQIUKFWjUqNF7me8kPDyc6tWr52jbZGZcuHCBtm3bSiDQQ8GCBWnfvn22x/r+kLxvx3tOkpqBEEIIqRkIIYSQYCCEEAIJBkIIIZBgIIQQAgkGQgghkGAghBAC+D+xSgf/m1754wAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Create an informative confusion matrix using seaborn\n",
    "#Define variables for Information inside each box\n",
    "category = ['True Neg', 'False Pos', 'False Neg', 'True Pos']\n",
    "counts = ['{0:0.0f}'.format(value) for value in confusion_matrix.flatten()]\n",
    "percentages = ['{0:.1%}'.format(value) for value in confusion_matrix.flatten()/np.sum(confusion_matrix)]\n",
    "labels = [f'{v1}\\n{v2}\\n{v3}' for v1, v2, v3 in zip(category, counts, percentages)]\n",
    "#Convert all of the labels to an array and reshape\n",
    "labels = np.asarray(labels).reshape(2,2)\n",
    "#Change the 0,1 to no stroke and stroke for labels\n",
    "tick_labels = ['No Stroke', 'Stroke']\n",
    "#Calculate all of the values to be displayed underneath the matrix as xlabel. Calculate precision, recall, F1 for both\n",
    "#stroke and no stroke\n",
    "accuracy  = np.trace(confusion_matrix) / float(np.sum(confusion_matrix))\n",
    "precision_stroke = confusion_matrix[1,1] / sum(confusion_matrix[:,1])\n",
    "recall_stroke    = confusion_matrix[1,1] / sum(confusion_matrix[1,:])\n",
    "f1_score_stroke  = 2*precision_stroke*recall_stroke / (precision_stroke + recall_stroke)\n",
    "precision = confusion_matrix[0,0] / sum(confusion_matrix[:,0])\n",
    "recall    = confusion_matrix[0,0] / sum(confusion_matrix[0,:])\n",
    "f1_score  = 2*precision*recall / (precision + recall)\n",
    "#Set up the text for the x label display underneath the matrix\n",
    "stats_text = \"\\n\\nAccuracy={:0.2f}\\nPrecision (stroke)={:0.2f}, Precision (no stroke)={:0.2f}\\nRecall (stroke)={:0.2f}, Recall (no stroke)={:0.2f}\\nF1 Score (stroke)={:0.2f}, F1 Score (no stroke)={:0.2f} \".format(accuracy,precision_stroke, precision, recall_stroke, recall, f1_score_stroke, f1_score)\n",
    "#Make the seaborn heatmap\n",
    "sns.heatmap(confusion_matrix, annot = labels, fmt = '', xticklabels = tick_labels, yticklabels = tick_labels)\n",
    "plt.xlabel(stats_text, fontsize = 14)\n",
    "plt.title(\"KNN Neighbor\", fontsize = 14)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Acc: 0.933\n"
     ]
    }
   ],
   "source": [
    "print('Test Acc: %.3f' % knn_grid.score(X_test_scaled, y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "    negative       0.93      1.00      0.97      1008\n",
      "    positive       0.00      0.00      0.00        71\n",
      "\n",
      "    accuracy                           0.93      1079\n",
      "   macro avg       0.47      0.50      0.48      1079\n",
      "weighted avg       0.87      0.93      0.90      1079\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Calculate classification report\n",
    "from sklearn.metrics import classification_report\n",
    "print(classification_report(y_test, predictions,\n",
    "                            target_names=[\"negative\", \"positive\"]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "# #Confusion matrix to see correct and incorrect predictions\n",
    "# from sklearn.metrics import confusion_matrix\n",
    "# confusion_matrix = confusion_matrix(y_test, predictions)\n",
    "# print(confusion_matrix)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "k=7 Test Acc: 0.933\n"
     ]
    }
   ],
   "source": [
    "# Note that k: 13 seems to be the best choice for this dataset\n",
    "knn = KNeighborsClassifier(n_neighbors=7)\n",
    "knn.fit(X_train_scaled, y_train)\n",
    "print('k=7 Test Acc: %.3f' % knn.score(X_test_scaled, y_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Neural Networks with Keras"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set the seed value for the notebook so the results are reproducible\n",
    "from numpy.random import seed\n",
    "seed(42)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Data Processing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.utils import to_categorical"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1., 0.],\n",
       "       [1., 0.],\n",
       "       [1., 0.],\n",
       "       ...,\n",
       "       [1., 0.],\n",
       "       [1., 0.],\n",
       "       [1., 0.]], dtype=float32)"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# One-hot encoding\n",
    "y_train_categorical = to_categorical(y_train)\n",
    "y_test_categorical = to_categorical(y_test)\n",
    "y_train_categorical"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Defining our Model Architecture (the layers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.models import Sequential\n",
    "\n",
    "model = Sequential()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.layers import Dense\n",
    "number_inputs = 10\n",
    "number_hidden_nodes = 4\n",
    "model.add(Dense(units=number_hidden_nodes,\n",
    "                activation='relu', input_dim=number_inputs))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "number_classes = 2\n",
    "model.add(Dense(units=number_classes, activation='softmax'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Model Summary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense (Dense)                (None, 4)                 44        \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 2)                 10        \n",
      "=================================================================\n",
      "Total params: 54\n",
      "Trainable params: 54\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Compile the Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(optimizer='adam',\n",
    "              loss='categorical_crossentropy',\n",
    "              metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Training the Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1000\n",
      "102/102 - 0s - loss: 0.4844 - accuracy: 0.8315\n",
      "Epoch 2/1000\n",
      "102/102 - 0s - loss: 0.3402 - accuracy: 0.9057\n",
      "Epoch 3/1000\n",
      "102/102 - 0s - loss: 0.2746 - accuracy: 0.9242\n",
      "Epoch 4/1000\n",
      "102/102 - 0s - loss: 0.2429 - accuracy: 0.9323\n",
      "Epoch 5/1000\n",
      "102/102 - 0s - loss: 0.2257 - accuracy: 0.9382\n",
      "Epoch 6/1000\n",
      "102/102 - 0s - loss: 0.2158 - accuracy: 0.9412\n",
      "Epoch 7/1000\n",
      "102/102 - 0s - loss: 0.2093 - accuracy: 0.9431\n",
      "Epoch 8/1000\n",
      "102/102 - 0s - loss: 0.2046 - accuracy: 0.9437\n",
      "Epoch 9/1000\n",
      "102/102 - 0s - loss: 0.2009 - accuracy: 0.9447\n",
      "Epoch 10/1000\n",
      "102/102 - 0s - loss: 0.1977 - accuracy: 0.9453\n",
      "Epoch 11/1000\n",
      "102/102 - 0s - loss: 0.1951 - accuracy: 0.9450\n",
      "Epoch 12/1000\n",
      "102/102 - 0s - loss: 0.1930 - accuracy: 0.9453\n",
      "Epoch 13/1000\n",
      "102/102 - 0s - loss: 0.1910 - accuracy: 0.9456\n",
      "Epoch 14/1000\n",
      "102/102 - 0s - loss: 0.1893 - accuracy: 0.9456\n",
      "Epoch 15/1000\n",
      "102/102 - 0s - loss: 0.1874 - accuracy: 0.9456\n",
      "Epoch 16/1000\n",
      "102/102 - 0s - loss: 0.1862 - accuracy: 0.9456\n",
      "Epoch 17/1000\n",
      "102/102 - 0s - loss: 0.1849 - accuracy: 0.9456\n",
      "Epoch 18/1000\n",
      "102/102 - 0s - loss: 0.1838 - accuracy: 0.9456\n",
      "Epoch 19/1000\n",
      "102/102 - 0s - loss: 0.1827 - accuracy: 0.9456\n",
      "Epoch 20/1000\n",
      "102/102 - 0s - loss: 0.1818 - accuracy: 0.9456\n",
      "Epoch 21/1000\n",
      "102/102 - 0s - loss: 0.1810 - accuracy: 0.9456\n",
      "Epoch 22/1000\n",
      "102/102 - 0s - loss: 0.1803 - accuracy: 0.9456\n",
      "Epoch 23/1000\n",
      "102/102 - 0s - loss: 0.1795 - accuracy: 0.9456\n",
      "Epoch 24/1000\n",
      "102/102 - 0s - loss: 0.1788 - accuracy: 0.9456\n",
      "Epoch 25/1000\n",
      "102/102 - 0s - loss: 0.1783 - accuracy: 0.9456\n",
      "Epoch 26/1000\n",
      "102/102 - 0s - loss: 0.1778 - accuracy: 0.9456\n",
      "Epoch 27/1000\n",
      "102/102 - 0s - loss: 0.1772 - accuracy: 0.9456\n",
      "Epoch 28/1000\n",
      "102/102 - 0s - loss: 0.1769 - accuracy: 0.9456\n",
      "Epoch 29/1000\n",
      "102/102 - 0s - loss: 0.1764 - accuracy: 0.9456\n",
      "Epoch 30/1000\n",
      "102/102 - 0s - loss: 0.1760 - accuracy: 0.9456\n",
      "Epoch 31/1000\n",
      "102/102 - 0s - loss: 0.1758 - accuracy: 0.9456\n",
      "Epoch 32/1000\n",
      "102/102 - 0s - loss: 0.1754 - accuracy: 0.9456\n",
      "Epoch 33/1000\n",
      "102/102 - 0s - loss: 0.1751 - accuracy: 0.9456\n",
      "Epoch 34/1000\n",
      "102/102 - 0s - loss: 0.1749 - accuracy: 0.9456\n",
      "Epoch 35/1000\n",
      "102/102 - 0s - loss: 0.1746 - accuracy: 0.9456\n",
      "Epoch 36/1000\n",
      "102/102 - 0s - loss: 0.1745 - accuracy: 0.9456\n",
      "Epoch 37/1000\n",
      "102/102 - 0s - loss: 0.1741 - accuracy: 0.9456\n",
      "Epoch 38/1000\n",
      "102/102 - 0s - loss: 0.1740 - accuracy: 0.9456\n",
      "Epoch 39/1000\n",
      "102/102 - 0s - loss: 0.1738 - accuracy: 0.9456\n",
      "Epoch 40/1000\n",
      "102/102 - 0s - loss: 0.1736 - accuracy: 0.9456\n",
      "Epoch 41/1000\n",
      "102/102 - 0s - loss: 0.1734 - accuracy: 0.9456\n",
      "Epoch 42/1000\n",
      "102/102 - 0s - loss: 0.1732 - accuracy: 0.9456\n",
      "Epoch 43/1000\n",
      "102/102 - 0s - loss: 0.1731 - accuracy: 0.9456\n",
      "Epoch 44/1000\n",
      "102/102 - 0s - loss: 0.1729 - accuracy: 0.9456\n",
      "Epoch 45/1000\n",
      "102/102 - 0s - loss: 0.1728 - accuracy: 0.9456\n",
      "Epoch 46/1000\n",
      "102/102 - 0s - loss: 0.1728 - accuracy: 0.9456\n",
      "Epoch 47/1000\n",
      "102/102 - 0s - loss: 0.1726 - accuracy: 0.9456\n",
      "Epoch 48/1000\n",
      "102/102 - 0s - loss: 0.1725 - accuracy: 0.9456\n",
      "Epoch 49/1000\n",
      "102/102 - 0s - loss: 0.1724 - accuracy: 0.9456\n",
      "Epoch 50/1000\n",
      "102/102 - 0s - loss: 0.1723 - accuracy: 0.9456\n",
      "Epoch 51/1000\n",
      "102/102 - 0s - loss: 0.1722 - accuracy: 0.9456\n",
      "Epoch 52/1000\n",
      "102/102 - 0s - loss: 0.1720 - accuracy: 0.9456\n",
      "Epoch 53/1000\n",
      "102/102 - 0s - loss: 0.1719 - accuracy: 0.9456\n",
      "Epoch 54/1000\n",
      "102/102 - 0s - loss: 0.1718 - accuracy: 0.9456\n",
      "Epoch 55/1000\n",
      "102/102 - 0s - loss: 0.1718 - accuracy: 0.9456\n",
      "Epoch 56/1000\n",
      "102/102 - 0s - loss: 0.1717 - accuracy: 0.9459\n",
      "Epoch 57/1000\n",
      "102/102 - 0s - loss: 0.1716 - accuracy: 0.9456\n",
      "Epoch 58/1000\n",
      "102/102 - 0s - loss: 0.1715 - accuracy: 0.9456\n",
      "Epoch 59/1000\n",
      "102/102 - 0s - loss: 0.1714 - accuracy: 0.9456\n",
      "Epoch 60/1000\n",
      "102/102 - 0s - loss: 0.1714 - accuracy: 0.9456\n",
      "Epoch 61/1000\n",
      "102/102 - 0s - loss: 0.1713 - accuracy: 0.9456\n",
      "Epoch 62/1000\n",
      "102/102 - 0s - loss: 0.1712 - accuracy: 0.9456\n",
      "Epoch 63/1000\n",
      "102/102 - 0s - loss: 0.1712 - accuracy: 0.9456\n",
      "Epoch 64/1000\n",
      "102/102 - 0s - loss: 0.1712 - accuracy: 0.9462\n",
      "Epoch 65/1000\n",
      "102/102 - 0s - loss: 0.1710 - accuracy: 0.9459\n",
      "Epoch 66/1000\n",
      "102/102 - 0s - loss: 0.1710 - accuracy: 0.9456\n",
      "Epoch 67/1000\n",
      "102/102 - 0s - loss: 0.1710 - accuracy: 0.9459\n",
      "Epoch 68/1000\n",
      "102/102 - 0s - loss: 0.1709 - accuracy: 0.9456\n",
      "Epoch 69/1000\n",
      "102/102 - 0s - loss: 0.1709 - accuracy: 0.9456\n",
      "Epoch 70/1000\n",
      "102/102 - 0s - loss: 0.1708 - accuracy: 0.9456\n",
      "Epoch 71/1000\n",
      "102/102 - 0s - loss: 0.1708 - accuracy: 0.9456\n",
      "Epoch 72/1000\n",
      "102/102 - 0s - loss: 0.1707 - accuracy: 0.9456\n",
      "Epoch 73/1000\n",
      "102/102 - 0s - loss: 0.1706 - accuracy: 0.9456\n",
      "Epoch 74/1000\n",
      "102/102 - 0s - loss: 0.1706 - accuracy: 0.9456\n",
      "Epoch 75/1000\n",
      "102/102 - 0s - loss: 0.1706 - accuracy: 0.9456\n",
      "Epoch 76/1000\n",
      "102/102 - 0s - loss: 0.1706 - accuracy: 0.9456\n",
      "Epoch 77/1000\n",
      "102/102 - 0s - loss: 0.1705 - accuracy: 0.9456\n",
      "Epoch 78/1000\n",
      "102/102 - 0s - loss: 0.1704 - accuracy: 0.9456\n",
      "Epoch 79/1000\n",
      "102/102 - 0s - loss: 0.1704 - accuracy: 0.9456\n",
      "Epoch 80/1000\n",
      "102/102 - 0s - loss: 0.1703 - accuracy: 0.9456\n",
      "Epoch 81/1000\n",
      "102/102 - 0s - loss: 0.1703 - accuracy: 0.9456\n",
      "Epoch 82/1000\n",
      "102/102 - 0s - loss: 0.1703 - accuracy: 0.9456\n",
      "Epoch 83/1000\n",
      "102/102 - 0s - loss: 0.1702 - accuracy: 0.9456\n",
      "Epoch 84/1000\n",
      "102/102 - 0s - loss: 0.1702 - accuracy: 0.9456\n",
      "Epoch 85/1000\n",
      "102/102 - 0s - loss: 0.1702 - accuracy: 0.9456\n",
      "Epoch 86/1000\n",
      "102/102 - 0s - loss: 0.1702 - accuracy: 0.9456\n",
      "Epoch 87/1000\n",
      "102/102 - 0s - loss: 0.1702 - accuracy: 0.9456\n",
      "Epoch 88/1000\n",
      "102/102 - 0s - loss: 0.1701 - accuracy: 0.9456\n",
      "Epoch 89/1000\n",
      "102/102 - 0s - loss: 0.1701 - accuracy: 0.9456\n",
      "Epoch 90/1000\n",
      "102/102 - 0s - loss: 0.1701 - accuracy: 0.9456\n",
      "Epoch 91/1000\n",
      "102/102 - 0s - loss: 0.1700 - accuracy: 0.9456\n",
      "Epoch 92/1000\n",
      "102/102 - 0s - loss: 0.1700 - accuracy: 0.9456\n",
      "Epoch 93/1000\n",
      "102/102 - 0s - loss: 0.1700 - accuracy: 0.9456\n",
      "Epoch 94/1000\n",
      "102/102 - 0s - loss: 0.1700 - accuracy: 0.9456\n",
      "Epoch 95/1000\n",
      "102/102 - 0s - loss: 0.1700 - accuracy: 0.9456\n",
      "Epoch 96/1000\n",
      "102/102 - 0s - loss: 0.1700 - accuracy: 0.9456\n",
      "Epoch 97/1000\n",
      "102/102 - 0s - loss: 0.1699 - accuracy: 0.9456\n",
      "Epoch 98/1000\n",
      "102/102 - 0s - loss: 0.1703 - accuracy: 0.9459\n",
      "Epoch 99/1000\n",
      "102/102 - 0s - loss: 0.1700 - accuracy: 0.9456\n",
      "Epoch 100/1000\n",
      "102/102 - 0s - loss: 0.1699 - accuracy: 0.9456\n",
      "Epoch 101/1000\n",
      "102/102 - 0s - loss: 0.1698 - accuracy: 0.9456\n",
      "Epoch 102/1000\n",
      "102/102 - 0s - loss: 0.1699 - accuracy: 0.9459\n",
      "Epoch 103/1000\n",
      "102/102 - 0s - loss: 0.1698 - accuracy: 0.9459\n",
      "Epoch 104/1000\n",
      "102/102 - 0s - loss: 0.1698 - accuracy: 0.9459\n",
      "Epoch 105/1000\n",
      "102/102 - 0s - loss: 0.1697 - accuracy: 0.9459\n",
      "Epoch 106/1000\n",
      "102/102 - 0s - loss: 0.1698 - accuracy: 0.9459\n",
      "Epoch 107/1000\n",
      "102/102 - 0s - loss: 0.1697 - accuracy: 0.9456\n",
      "Epoch 108/1000\n",
      "102/102 - 0s - loss: 0.1697 - accuracy: 0.9456\n",
      "Epoch 109/1000\n",
      "102/102 - 0s - loss: 0.1698 - accuracy: 0.9453\n",
      "Epoch 110/1000\n",
      "102/102 - 0s - loss: 0.1698 - accuracy: 0.9456\n",
      "Epoch 111/1000\n",
      "102/102 - 0s - loss: 0.1697 - accuracy: 0.9456\n",
      "Epoch 112/1000\n",
      "102/102 - 0s - loss: 0.1698 - accuracy: 0.9456\n",
      "Epoch 113/1000\n",
      "102/102 - 0s - loss: 0.1697 - accuracy: 0.9456\n",
      "Epoch 114/1000\n",
      "102/102 - 0s - loss: 0.1697 - accuracy: 0.9456\n",
      "Epoch 115/1000\n",
      "102/102 - 0s - loss: 0.1697 - accuracy: 0.9456\n",
      "Epoch 116/1000\n",
      "102/102 - 0s - loss: 0.1696 - accuracy: 0.9453\n",
      "Epoch 117/1000\n",
      "102/102 - 0s - loss: 0.1697 - accuracy: 0.9456\n",
      "Epoch 118/1000\n",
      "102/102 - 0s - loss: 0.1696 - accuracy: 0.9456\n",
      "Epoch 119/1000\n",
      "102/102 - 0s - loss: 0.1696 - accuracy: 0.9456\n",
      "Epoch 120/1000\n",
      "102/102 - 0s - loss: 0.1696 - accuracy: 0.9456\n",
      "Epoch 121/1000\n",
      "102/102 - 0s - loss: 0.1695 - accuracy: 0.9456\n",
      "Epoch 122/1000\n",
      "102/102 - 0s - loss: 0.1695 - accuracy: 0.9456\n",
      "Epoch 123/1000\n",
      "102/102 - 0s - loss: 0.1694 - accuracy: 0.9456\n",
      "Epoch 124/1000\n",
      "102/102 - 0s - loss: 0.1695 - accuracy: 0.9456\n",
      "Epoch 125/1000\n",
      "102/102 - 0s - loss: 0.1694 - accuracy: 0.9456\n",
      "Epoch 126/1000\n",
      "102/102 - 0s - loss: 0.1695 - accuracy: 0.9456\n",
      "Epoch 127/1000\n",
      "102/102 - 0s - loss: 0.1694 - accuracy: 0.9459\n",
      "Epoch 128/1000\n",
      "102/102 - 0s - loss: 0.1694 - accuracy: 0.9456\n",
      "Epoch 129/1000\n",
      "102/102 - 0s - loss: 0.1693 - accuracy: 0.9456\n",
      "Epoch 130/1000\n",
      "102/102 - 0s - loss: 0.1694 - accuracy: 0.9456\n",
      "Epoch 131/1000\n",
      "102/102 - 0s - loss: 0.1693 - accuracy: 0.9456\n",
      "Epoch 132/1000\n",
      "102/102 - 0s - loss: 0.1692 - accuracy: 0.9456\n",
      "Epoch 133/1000\n",
      "102/102 - 0s - loss: 0.1692 - accuracy: 0.9456\n",
      "Epoch 134/1000\n",
      "102/102 - 0s - loss: 0.1693 - accuracy: 0.9456\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 135/1000\n",
      "102/102 - 0s - loss: 0.1692 - accuracy: 0.9456\n",
      "Epoch 136/1000\n",
      "102/102 - 0s - loss: 0.1693 - accuracy: 0.9456\n",
      "Epoch 137/1000\n",
      "102/102 - 0s - loss: 0.1693 - accuracy: 0.9459\n",
      "Epoch 138/1000\n",
      "102/102 - 0s - loss: 0.1693 - accuracy: 0.9456\n",
      "Epoch 139/1000\n",
      "102/102 - 0s - loss: 0.1692 - accuracy: 0.9459\n",
      "Epoch 140/1000\n",
      "102/102 - 0s - loss: 0.1692 - accuracy: 0.9453\n",
      "Epoch 141/1000\n",
      "102/102 - 0s - loss: 0.1692 - accuracy: 0.9456\n",
      "Epoch 142/1000\n",
      "102/102 - 0s - loss: 0.1691 - accuracy: 0.9456\n",
      "Epoch 143/1000\n",
      "102/102 - 0s - loss: 0.1691 - accuracy: 0.9456\n",
      "Epoch 144/1000\n",
      "102/102 - 0s - loss: 0.1692 - accuracy: 0.9456\n",
      "Epoch 145/1000\n",
      "102/102 - 0s - loss: 0.1692 - accuracy: 0.9456\n",
      "Epoch 146/1000\n",
      "102/102 - 0s - loss: 0.1691 - accuracy: 0.9459\n",
      "Epoch 147/1000\n",
      "102/102 - 0s - loss: 0.1691 - accuracy: 0.9459\n",
      "Epoch 148/1000\n",
      "102/102 - 0s - loss: 0.1691 - accuracy: 0.9456\n",
      "Epoch 149/1000\n",
      "102/102 - 0s - loss: 0.1690 - accuracy: 0.9456\n",
      "Epoch 150/1000\n",
      "102/102 - 0s - loss: 0.1691 - accuracy: 0.9456\n",
      "Epoch 151/1000\n",
      "102/102 - 0s - loss: 0.1690 - accuracy: 0.9456\n",
      "Epoch 152/1000\n",
      "102/102 - 0s - loss: 0.1692 - accuracy: 0.9459\n",
      "Epoch 153/1000\n",
      "102/102 - 0s - loss: 0.1691 - accuracy: 0.9459\n",
      "Epoch 154/1000\n",
      "102/102 - 0s - loss: 0.1692 - accuracy: 0.9459\n",
      "Epoch 155/1000\n",
      "102/102 - 0s - loss: 0.1691 - accuracy: 0.9456\n",
      "Epoch 156/1000\n",
      "102/102 - 0s - loss: 0.1690 - accuracy: 0.9456\n",
      "Epoch 157/1000\n",
      "102/102 - 0s - loss: 0.1690 - accuracy: 0.9459\n",
      "Epoch 158/1000\n",
      "102/102 - 0s - loss: 0.1690 - accuracy: 0.9456\n",
      "Epoch 159/1000\n",
      "102/102 - 0s - loss: 0.1690 - accuracy: 0.9456\n",
      "Epoch 160/1000\n",
      "102/102 - 0s - loss: 0.1690 - accuracy: 0.9459\n",
      "Epoch 161/1000\n",
      "102/102 - 0s - loss: 0.1690 - accuracy: 0.9456\n",
      "Epoch 162/1000\n",
      "102/102 - 0s - loss: 0.1690 - accuracy: 0.9456\n",
      "Epoch 163/1000\n",
      "102/102 - 0s - loss: 0.1689 - accuracy: 0.9456\n",
      "Epoch 164/1000\n",
      "102/102 - 0s - loss: 0.1689 - accuracy: 0.9453\n",
      "Epoch 165/1000\n",
      "102/102 - 0s - loss: 0.1689 - accuracy: 0.9456\n",
      "Epoch 166/1000\n",
      "102/102 - 0s - loss: 0.1689 - accuracy: 0.9456\n",
      "Epoch 167/1000\n",
      "102/102 - 0s - loss: 0.1689 - accuracy: 0.9456\n",
      "Epoch 168/1000\n",
      "102/102 - 0s - loss: 0.1690 - accuracy: 0.9459\n",
      "Epoch 169/1000\n",
      "102/102 - 0s - loss: 0.1689 - accuracy: 0.9459\n",
      "Epoch 170/1000\n",
      "102/102 - 0s - loss: 0.1689 - accuracy: 0.9456\n",
      "Epoch 171/1000\n",
      "102/102 - 0s - loss: 0.1690 - accuracy: 0.9453\n",
      "Epoch 172/1000\n",
      "102/102 - 0s - loss: 0.1688 - accuracy: 0.9459\n",
      "Epoch 173/1000\n",
      "102/102 - 0s - loss: 0.1689 - accuracy: 0.9456\n",
      "Epoch 174/1000\n",
      "102/102 - 0s - loss: 0.1689 - accuracy: 0.9453\n",
      "Epoch 175/1000\n",
      "102/102 - 0s - loss: 0.1689 - accuracy: 0.9459\n",
      "Epoch 176/1000\n",
      "102/102 - 0s - loss: 0.1690 - accuracy: 0.9456\n",
      "Epoch 177/1000\n",
      "102/102 - 0s - loss: 0.1688 - accuracy: 0.9456\n",
      "Epoch 178/1000\n",
      "102/102 - 0s - loss: 0.1689 - accuracy: 0.9453\n",
      "Epoch 179/1000\n",
      "102/102 - 0s - loss: 0.1688 - accuracy: 0.9453\n",
      "Epoch 180/1000\n",
      "102/102 - 0s - loss: 0.1689 - accuracy: 0.9453\n",
      "Epoch 181/1000\n",
      "102/102 - 0s - loss: 0.1688 - accuracy: 0.9453\n",
      "Epoch 182/1000\n",
      "102/102 - 0s - loss: 0.1689 - accuracy: 0.9453\n",
      "Epoch 183/1000\n",
      "102/102 - 0s - loss: 0.1688 - accuracy: 0.9453\n",
      "Epoch 184/1000\n",
      "102/102 - 0s - loss: 0.1688 - accuracy: 0.9453\n",
      "Epoch 185/1000\n",
      "102/102 - 0s - loss: 0.1688 - accuracy: 0.9453\n",
      "Epoch 186/1000\n",
      "102/102 - 0s - loss: 0.1688 - accuracy: 0.9453\n",
      "Epoch 187/1000\n",
      "102/102 - 0s - loss: 0.1687 - accuracy: 0.9453\n",
      "Epoch 188/1000\n",
      "102/102 - 0s - loss: 0.1688 - accuracy: 0.9453\n",
      "Epoch 189/1000\n",
      "102/102 - 0s - loss: 0.1687 - accuracy: 0.9453\n",
      "Epoch 190/1000\n",
      "102/102 - 0s - loss: 0.1687 - accuracy: 0.9453\n",
      "Epoch 191/1000\n",
      "102/102 - 0s - loss: 0.1688 - accuracy: 0.9453\n",
      "Epoch 192/1000\n",
      "102/102 - 0s - loss: 0.1688 - accuracy: 0.9453\n",
      "Epoch 193/1000\n",
      "102/102 - 0s - loss: 0.1688 - accuracy: 0.9453\n",
      "Epoch 194/1000\n",
      "102/102 - 0s - loss: 0.1687 - accuracy: 0.9453\n",
      "Epoch 195/1000\n",
      "102/102 - 0s - loss: 0.1687 - accuracy: 0.9453\n",
      "Epoch 196/1000\n",
      "102/102 - 0s - loss: 0.1687 - accuracy: 0.9453\n",
      "Epoch 197/1000\n",
      "102/102 - 0s - loss: 0.1688 - accuracy: 0.9453\n",
      "Epoch 198/1000\n",
      "102/102 - 0s - loss: 0.1687 - accuracy: 0.9456\n",
      "Epoch 199/1000\n",
      "102/102 - 0s - loss: 0.1686 - accuracy: 0.9453\n",
      "Epoch 200/1000\n",
      "102/102 - 0s - loss: 0.1686 - accuracy: 0.9453\n",
      "Epoch 201/1000\n",
      "102/102 - 0s - loss: 0.1686 - accuracy: 0.9453\n",
      "Epoch 202/1000\n",
      "102/102 - 0s - loss: 0.1687 - accuracy: 0.9453\n",
      "Epoch 203/1000\n",
      "102/102 - 0s - loss: 0.1689 - accuracy: 0.9453\n",
      "Epoch 204/1000\n",
      "102/102 - 0s - loss: 0.1687 - accuracy: 0.9456\n",
      "Epoch 205/1000\n",
      "102/102 - 0s - loss: 0.1686 - accuracy: 0.9453\n",
      "Epoch 206/1000\n",
      "102/102 - 0s - loss: 0.1686 - accuracy: 0.9456\n",
      "Epoch 207/1000\n",
      "102/102 - 0s - loss: 0.1686 - accuracy: 0.9453\n",
      "Epoch 208/1000\n",
      "102/102 - 0s - loss: 0.1686 - accuracy: 0.9453\n",
      "Epoch 209/1000\n",
      "102/102 - 0s - loss: 0.1686 - accuracy: 0.9453\n",
      "Epoch 210/1000\n",
      "102/102 - 0s - loss: 0.1685 - accuracy: 0.9453\n",
      "Epoch 211/1000\n",
      "102/102 - 0s - loss: 0.1686 - accuracy: 0.9453\n",
      "Epoch 212/1000\n",
      "102/102 - 0s - loss: 0.1686 - accuracy: 0.9453\n",
      "Epoch 213/1000\n",
      "102/102 - 0s - loss: 0.1686 - accuracy: 0.9456\n",
      "Epoch 214/1000\n",
      "102/102 - 0s - loss: 0.1686 - accuracy: 0.9453\n",
      "Epoch 215/1000\n",
      "102/102 - 0s - loss: 0.1686 - accuracy: 0.9453\n",
      "Epoch 216/1000\n",
      "102/102 - 0s - loss: 0.1686 - accuracy: 0.9450\n",
      "Epoch 217/1000\n",
      "102/102 - 0s - loss: 0.1685 - accuracy: 0.9453\n",
      "Epoch 218/1000\n",
      "102/102 - 0s - loss: 0.1685 - accuracy: 0.9450\n",
      "Epoch 219/1000\n",
      "102/102 - 0s - loss: 0.1685 - accuracy: 0.9450\n",
      "Epoch 220/1000\n",
      "102/102 - 0s - loss: 0.1685 - accuracy: 0.9453\n",
      "Epoch 221/1000\n",
      "102/102 - 0s - loss: 0.1684 - accuracy: 0.9453\n",
      "Epoch 222/1000\n",
      "102/102 - 0s - loss: 0.1688 - accuracy: 0.9453\n",
      "Epoch 223/1000\n",
      "102/102 - 0s - loss: 0.1686 - accuracy: 0.9453\n",
      "Epoch 224/1000\n",
      "102/102 - 0s - loss: 0.1685 - accuracy: 0.9453\n",
      "Epoch 225/1000\n",
      "102/102 - 0s - loss: 0.1685 - accuracy: 0.9453\n",
      "Epoch 226/1000\n",
      "102/102 - 0s - loss: 0.1685 - accuracy: 0.9453\n",
      "Epoch 227/1000\n",
      "102/102 - 0s - loss: 0.1685 - accuracy: 0.9453\n",
      "Epoch 228/1000\n",
      "102/102 - 0s - loss: 0.1684 - accuracy: 0.9453\n",
      "Epoch 229/1000\n",
      "102/102 - 0s - loss: 0.1684 - accuracy: 0.9450\n",
      "Epoch 230/1000\n",
      "102/102 - 0s - loss: 0.1686 - accuracy: 0.9453\n",
      "Epoch 231/1000\n",
      "102/102 - 0s - loss: 0.1685 - accuracy: 0.9450\n",
      "Epoch 232/1000\n",
      "102/102 - 0s - loss: 0.1684 - accuracy: 0.9450\n",
      "Epoch 233/1000\n",
      "102/102 - 0s - loss: 0.1684 - accuracy: 0.9453\n",
      "Epoch 234/1000\n",
      "102/102 - 0s - loss: 0.1685 - accuracy: 0.9453\n",
      "Epoch 235/1000\n",
      "102/102 - 0s - loss: 0.1685 - accuracy: 0.9453\n",
      "Epoch 236/1000\n",
      "102/102 - 0s - loss: 0.1685 - accuracy: 0.9453\n",
      "Epoch 237/1000\n",
      "102/102 - 0s - loss: 0.1684 - accuracy: 0.9450\n",
      "Epoch 238/1000\n",
      "102/102 - 0s - loss: 0.1685 - accuracy: 0.9453\n",
      "Epoch 239/1000\n",
      "102/102 - 0s - loss: 0.1684 - accuracy: 0.9453\n",
      "Epoch 240/1000\n",
      "102/102 - 0s - loss: 0.1684 - accuracy: 0.9450\n",
      "Epoch 241/1000\n",
      "102/102 - 0s - loss: 0.1684 - accuracy: 0.9450\n",
      "Epoch 242/1000\n",
      "102/102 - 0s - loss: 0.1685 - accuracy: 0.9453\n",
      "Epoch 243/1000\n",
      "102/102 - 0s - loss: 0.1685 - accuracy: 0.9453\n",
      "Epoch 244/1000\n",
      "102/102 - 0s - loss: 0.1683 - accuracy: 0.9453\n",
      "Epoch 245/1000\n",
      "102/102 - 0s - loss: 0.1683 - accuracy: 0.9453\n",
      "Epoch 246/1000\n",
      "102/102 - 0s - loss: 0.1684 - accuracy: 0.9453\n",
      "Epoch 247/1000\n",
      "102/102 - 0s - loss: 0.1684 - accuracy: 0.9453\n",
      "Epoch 248/1000\n",
      "102/102 - 0s - loss: 0.1683 - accuracy: 0.9453\n",
      "Epoch 249/1000\n",
      "102/102 - 0s - loss: 0.1683 - accuracy: 0.9450\n",
      "Epoch 250/1000\n",
      "102/102 - 0s - loss: 0.1683 - accuracy: 0.9453\n",
      "Epoch 251/1000\n",
      "102/102 - 0s - loss: 0.1683 - accuracy: 0.9450\n",
      "Epoch 252/1000\n",
      "102/102 - 0s - loss: 0.1683 - accuracy: 0.9450\n",
      "Epoch 253/1000\n",
      "102/102 - 0s - loss: 0.1683 - accuracy: 0.9450\n",
      "Epoch 254/1000\n",
      "102/102 - 0s - loss: 0.1683 - accuracy: 0.9450\n",
      "Epoch 255/1000\n",
      "102/102 - 0s - loss: 0.1683 - accuracy: 0.9450\n",
      "Epoch 256/1000\n",
      "102/102 - 0s - loss: 0.1683 - accuracy: 0.9453\n",
      "Epoch 257/1000\n",
      "102/102 - 0s - loss: 0.1684 - accuracy: 0.9450\n",
      "Epoch 258/1000\n",
      "102/102 - 0s - loss: 0.1683 - accuracy: 0.9453\n",
      "Epoch 259/1000\n",
      "102/102 - 0s - loss: 0.1683 - accuracy: 0.9453\n",
      "Epoch 260/1000\n",
      "102/102 - 0s - loss: 0.1683 - accuracy: 0.9450\n",
      "Epoch 261/1000\n",
      "102/102 - 0s - loss: 0.1682 - accuracy: 0.9453\n",
      "Epoch 262/1000\n",
      "102/102 - 0s - loss: 0.1682 - accuracy: 0.9453\n",
      "Epoch 263/1000\n",
      "102/102 - 0s - loss: 0.1683 - accuracy: 0.9450\n",
      "Epoch 264/1000\n",
      "102/102 - 0s - loss: 0.1683 - accuracy: 0.9453\n",
      "Epoch 265/1000\n",
      "102/102 - 0s - loss: 0.1682 - accuracy: 0.9453\n",
      "Epoch 266/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9453\n",
      "Epoch 267/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "102/102 - 0s - loss: 0.1682 - accuracy: 0.9453\n",
      "Epoch 268/1000\n",
      "102/102 - 0s - loss: 0.1682 - accuracy: 0.9450\n",
      "Epoch 269/1000\n",
      "102/102 - 0s - loss: 0.1682 - accuracy: 0.9453\n",
      "Epoch 270/1000\n",
      "102/102 - 0s - loss: 0.1682 - accuracy: 0.9453\n",
      "Epoch 271/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9450\n",
      "Epoch 272/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9453\n",
      "Epoch 273/1000\n",
      "102/102 - 0s - loss: 0.1682 - accuracy: 0.9453\n",
      "Epoch 274/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9453\n",
      "Epoch 275/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9453\n",
      "Epoch 276/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9453\n",
      "Epoch 277/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9453\n",
      "Epoch 278/1000\n",
      "102/102 - 0s - loss: 0.1684 - accuracy: 0.9450\n",
      "Epoch 279/1000\n",
      "102/102 - 0s - loss: 0.1683 - accuracy: 0.9453\n",
      "Epoch 280/1000\n",
      "102/102 - 0s - loss: 0.1682 - accuracy: 0.9453\n",
      "Epoch 281/1000\n",
      "102/102 - 0s - loss: 0.1682 - accuracy: 0.9453\n",
      "Epoch 282/1000\n",
      "102/102 - 0s - loss: 0.1682 - accuracy: 0.9453\n",
      "Epoch 283/1000\n",
      "102/102 - 0s - loss: 0.1682 - accuracy: 0.9450\n",
      "Epoch 284/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9450\n",
      "Epoch 285/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9450\n",
      "Epoch 286/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9450\n",
      "Epoch 287/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9450\n",
      "Epoch 288/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9453\n",
      "Epoch 289/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9450\n",
      "Epoch 290/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9453\n",
      "Epoch 291/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9453\n",
      "Epoch 292/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9453\n",
      "Epoch 293/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9450\n",
      "Epoch 294/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9450\n",
      "Epoch 295/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9453\n",
      "Epoch 296/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9453\n",
      "Epoch 297/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9450\n",
      "Epoch 298/1000\n",
      "102/102 - 0s - loss: 0.1684 - accuracy: 0.9453\n",
      "Epoch 299/1000\n",
      "102/102 - 0s - loss: 0.1683 - accuracy: 0.9453\n",
      "Epoch 300/1000\n",
      "102/102 - 0s - loss: 0.1683 - accuracy: 0.9450\n",
      "Epoch 301/1000\n",
      "102/102 - 0s - loss: 0.1683 - accuracy: 0.9450\n",
      "Epoch 302/1000\n",
      "102/102 - 0s - loss: 0.1682 - accuracy: 0.9453\n",
      "Epoch 303/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9450\n",
      "Epoch 304/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9453\n",
      "Epoch 305/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9450\n",
      "Epoch 306/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9453\n",
      "Epoch 307/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9450\n",
      "Epoch 308/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9450\n",
      "Epoch 309/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9450\n",
      "Epoch 310/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 311/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 312/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9453\n",
      "Epoch 313/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9450\n",
      "Epoch 314/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9453\n",
      "Epoch 315/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9450\n",
      "Epoch 316/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9453\n",
      "Epoch 317/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9453\n",
      "Epoch 318/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9450\n",
      "Epoch 319/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 320/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9450\n",
      "Epoch 321/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9450\n",
      "Epoch 322/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9453\n",
      "Epoch 323/1000\n",
      "102/102 - 0s - loss: 0.1683 - accuracy: 0.9453\n",
      "Epoch 324/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9450\n",
      "Epoch 325/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9453\n",
      "Epoch 326/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 327/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 328/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9450\n",
      "Epoch 329/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 330/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 331/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 332/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9450\n",
      "Epoch 333/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 334/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 335/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 336/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 337/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9453\n",
      "Epoch 338/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9450\n",
      "Epoch 339/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9453\n",
      "Epoch 340/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9453\n",
      "Epoch 341/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 342/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9453\n",
      "Epoch 343/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 344/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9453\n",
      "Epoch 345/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 346/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9447\n",
      "Epoch 347/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 348/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 349/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9453\n",
      "Epoch 350/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9453\n",
      "Epoch 351/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 352/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 353/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9453\n",
      "Epoch 354/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9447\n",
      "Epoch 355/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9453\n",
      "Epoch 356/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 357/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 358/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 359/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 360/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 361/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 362/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 363/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9453\n",
      "Epoch 364/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 365/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9450\n",
      "Epoch 366/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 367/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9450\n",
      "Epoch 368/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9453\n",
      "Epoch 369/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 370/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9453\n",
      "Epoch 371/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 372/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 373/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 374/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 375/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 376/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 377/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 378/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 379/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 380/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 381/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 382/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 383/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9453\n",
      "Epoch 384/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9447\n",
      "Epoch 385/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 386/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 387/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 388/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 389/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 390/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 391/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 392/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 393/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 394/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 395/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 396/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 397/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 398/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 399/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 400/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 401/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 402/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9453\n",
      "Epoch 403/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9453\n",
      "Epoch 404/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9453\n",
      "Epoch 405/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 406/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 407/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 408/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9453\n",
      "Epoch 409/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9453\n",
      "Epoch 410/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9453\n",
      "Epoch 411/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 412/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 413/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9453\n",
      "Epoch 414/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9453\n",
      "Epoch 415/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9453\n",
      "Epoch 416/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 417/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 418/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 419/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 420/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 421/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 422/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 423/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9453\n",
      "Epoch 424/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 425/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 426/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 427/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9453\n",
      "Epoch 428/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 429/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 430/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 431/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 432/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9453\n",
      "Epoch 433/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 434/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 435/1000\n",
      "102/102 - 0s - loss: 0.1681 - accuracy: 0.9450\n",
      "Epoch 436/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9453\n",
      "Epoch 437/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 438/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 439/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 440/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9453\n",
      "Epoch 441/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 442/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 443/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 444/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 445/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 446/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 447/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 448/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 449/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 450/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9453\n",
      "Epoch 451/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 452/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 453/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 454/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 455/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 456/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 457/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 458/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 459/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9453\n",
      "Epoch 460/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 461/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 462/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 463/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 464/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 465/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 466/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 467/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 468/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 469/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 470/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 471/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 472/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 473/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 474/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 475/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 476/1000\n",
      "102/102 - 0s - loss: 0.1680 - accuracy: 0.9450\n",
      "Epoch 477/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9453\n",
      "Epoch 478/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 479/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 480/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 481/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 482/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9453\n",
      "Epoch 483/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 484/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9447\n",
      "Epoch 485/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 486/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 487/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 488/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 489/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 490/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 491/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 492/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 493/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 494/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 495/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 496/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 497/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 498/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 499/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 500/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 501/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 502/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 503/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 504/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9453\n",
      "Epoch 505/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 506/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 507/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 508/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9447\n",
      "Epoch 509/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 510/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 511/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 512/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 513/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 514/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 515/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 516/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 517/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 518/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 519/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 520/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 521/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9447\n",
      "Epoch 522/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 523/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 524/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 525/1000\n",
      "102/102 - 0s - loss: 0.1675 - accuracy: 0.9450\n",
      "Epoch 526/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 527/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9450\n",
      "Epoch 528/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 529/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9453\n",
      "Epoch 530/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 531/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 532/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 533/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 534/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 535/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 536/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 537/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 538/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9453\n",
      "Epoch 539/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9450\n",
      "Epoch 540/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 541/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 542/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 543/1000\n",
      "102/102 - 0s - loss: 0.1675 - accuracy: 0.9450\n",
      "Epoch 544/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 545/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 546/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9450\n",
      "Epoch 547/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 548/1000\n",
      "102/102 - 0s - loss: 0.1679 - accuracy: 0.9456\n",
      "Epoch 549/1000\n",
      "102/102 - 0s - loss: 0.1677 - accuracy: 0.9447\n",
      "Epoch 550/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 551/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 552/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9450\n",
      "Epoch 553/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9453\n",
      "Epoch 554/1000\n",
      "102/102 - 0s - loss: 0.1675 - accuracy: 0.9450\n",
      "Epoch 555/1000\n",
      "102/102 - 0s - loss: 0.1675 - accuracy: 0.9450\n",
      "Epoch 556/1000\n",
      "102/102 - 0s - loss: 0.1675 - accuracy: 0.9447\n",
      "Epoch 557/1000\n",
      "102/102 - 0s - loss: 0.1674 - accuracy: 0.9453\n",
      "Epoch 558/1000\n",
      "102/102 - 0s - loss: 0.1675 - accuracy: 0.9450\n",
      "Epoch 559/1000\n",
      "102/102 - 0s - loss: 0.1675 - accuracy: 0.9450\n",
      "Epoch 560/1000\n",
      "102/102 - 0s - loss: 0.1674 - accuracy: 0.9450\n",
      "Epoch 561/1000\n",
      "102/102 - 0s - loss: 0.1674 - accuracy: 0.9447\n",
      "Epoch 562/1000\n",
      "102/102 - 0s - loss: 0.1675 - accuracy: 0.9453\n",
      "Epoch 563/1000\n",
      "102/102 - 0s - loss: 0.1675 - accuracy: 0.9447\n",
      "Epoch 564/1000\n",
      "102/102 - 0s - loss: 0.1675 - accuracy: 0.9450\n",
      "Epoch 565/1000\n",
      "102/102 - 0s - loss: 0.1674 - accuracy: 0.9453\n",
      "Epoch 566/1000\n",
      "102/102 - 0s - loss: 0.1674 - accuracy: 0.9447\n",
      "Epoch 567/1000\n",
      "102/102 - 0s - loss: 0.1678 - accuracy: 0.9456\n",
      "Epoch 568/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9459\n",
      "Epoch 569/1000\n",
      "102/102 - 0s - loss: 0.1675 - accuracy: 0.9459\n",
      "Epoch 570/1000\n",
      "102/102 - 0s - loss: 0.1676 - accuracy: 0.9456\n",
      "Epoch 571/1000\n",
      "102/102 - 0s - loss: 0.1674 - accuracy: 0.9453\n",
      "Epoch 572/1000\n",
      "102/102 - 0s - loss: 0.1674 - accuracy: 0.9450\n",
      "Epoch 573/1000\n",
      "102/102 - 0s - loss: 0.1673 - accuracy: 0.9450\n",
      "Epoch 574/1000\n",
      "102/102 - 0s - loss: 0.1674 - accuracy: 0.9453\n",
      "Epoch 575/1000\n",
      "102/102 - 0s - loss: 0.1674 - accuracy: 0.9450\n",
      "Epoch 576/1000\n",
      "102/102 - 0s - loss: 0.1675 - accuracy: 0.9456\n",
      "Epoch 577/1000\n",
      "102/102 - 0s - loss: 0.1675 - accuracy: 0.9453\n",
      "Epoch 578/1000\n",
      "102/102 - 0s - loss: 0.1674 - accuracy: 0.9453\n",
      "Epoch 579/1000\n",
      "102/102 - 0s - loss: 0.1674 - accuracy: 0.9453\n",
      "Epoch 580/1000\n",
      "102/102 - 0s - loss: 0.1674 - accuracy: 0.9456\n",
      "Epoch 581/1000\n",
      "102/102 - 0s - loss: 0.1674 - accuracy: 0.9456\n",
      "Epoch 582/1000\n",
      "102/102 - 0s - loss: 0.1674 - accuracy: 0.9459\n",
      "Epoch 583/1000\n",
      "102/102 - 0s - loss: 0.1673 - accuracy: 0.9456\n",
      "Epoch 584/1000\n",
      "102/102 - 0s - loss: 0.1675 - accuracy: 0.9459\n",
      "Epoch 585/1000\n",
      "102/102 - 0s - loss: 0.1674 - accuracy: 0.9459\n",
      "Epoch 586/1000\n",
      "102/102 - 0s - loss: 0.1673 - accuracy: 0.9456\n",
      "Epoch 587/1000\n",
      "102/102 - 0s - loss: 0.1673 - accuracy: 0.9459\n",
      "Epoch 588/1000\n",
      "102/102 - 0s - loss: 0.1673 - accuracy: 0.9459\n",
      "Epoch 589/1000\n",
      "102/102 - 0s - loss: 0.1673 - accuracy: 0.9459\n",
      "Epoch 590/1000\n",
      "102/102 - 0s - loss: 0.1673 - accuracy: 0.9453\n",
      "Epoch 591/1000\n",
      "102/102 - 0s - loss: 0.1672 - accuracy: 0.9456\n",
      "Epoch 592/1000\n",
      "102/102 - 0s - loss: 0.1673 - accuracy: 0.9456\n",
      "Epoch 593/1000\n",
      "102/102 - 0s - loss: 0.1673 - accuracy: 0.9456\n",
      "Epoch 594/1000\n",
      "102/102 - 0s - loss: 0.1672 - accuracy: 0.9453\n",
      "Epoch 595/1000\n",
      "102/102 - 0s - loss: 0.1672 - accuracy: 0.9453\n",
      "Epoch 596/1000\n",
      "102/102 - 0s - loss: 0.1673 - accuracy: 0.9456\n",
      "Epoch 597/1000\n",
      "102/102 - 0s - loss: 0.1672 - accuracy: 0.9453\n",
      "Epoch 598/1000\n",
      "102/102 - 0s - loss: 0.1672 - accuracy: 0.9450\n",
      "Epoch 599/1000\n",
      "102/102 - 0s - loss: 0.1673 - accuracy: 0.9453\n",
      "Epoch 600/1000\n",
      "102/102 - 0s - loss: 0.1673 - accuracy: 0.9456\n",
      "Epoch 601/1000\n",
      "102/102 - 0s - loss: 0.1672 - accuracy: 0.9459\n",
      "Epoch 602/1000\n",
      "102/102 - 0s - loss: 0.1672 - accuracy: 0.9453\n",
      "Epoch 603/1000\n",
      "102/102 - 0s - loss: 0.1671 - accuracy: 0.9456\n",
      "Epoch 604/1000\n",
      "102/102 - 0s - loss: 0.1672 - accuracy: 0.9459\n",
      "Epoch 605/1000\n",
      "102/102 - 0s - loss: 0.1672 - accuracy: 0.9459\n",
      "Epoch 606/1000\n",
      "102/102 - 0s - loss: 0.1672 - accuracy: 0.9456\n",
      "Epoch 607/1000\n",
      "102/102 - 0s - loss: 0.1671 - accuracy: 0.9453\n",
      "Epoch 608/1000\n",
      "102/102 - 0s - loss: 0.1672 - accuracy: 0.9462\n",
      "Epoch 609/1000\n",
      "102/102 - 0s - loss: 0.1672 - accuracy: 0.9459\n",
      "Epoch 610/1000\n",
      "102/102 - 0s - loss: 0.1672 - accuracy: 0.9456\n",
      "Epoch 611/1000\n",
      "102/102 - 0s - loss: 0.1671 - accuracy: 0.9456\n",
      "Epoch 612/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9459\n",
      "Epoch 613/1000\n",
      "102/102 - 0s - loss: 0.1672 - accuracy: 0.9456\n",
      "Epoch 614/1000\n",
      "102/102 - 0s - loss: 0.1672 - accuracy: 0.9459\n",
      "Epoch 615/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9456\n",
      "Epoch 616/1000\n",
      "102/102 - 0s - loss: 0.1671 - accuracy: 0.9459\n",
      "Epoch 617/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9456\n",
      "Epoch 618/1000\n",
      "102/102 - 0s - loss: 0.1671 - accuracy: 0.9459\n",
      "Epoch 619/1000\n",
      "102/102 - 0s - loss: 0.1671 - accuracy: 0.9459\n",
      "Epoch 620/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9459\n",
      "Epoch 621/1000\n",
      "102/102 - 0s - loss: 0.1673 - accuracy: 0.9456\n",
      "Epoch 622/1000\n",
      "102/102 - 0s - loss: 0.1672 - accuracy: 0.9456\n",
      "Epoch 623/1000\n",
      "102/102 - 0s - loss: 0.1671 - accuracy: 0.9456\n",
      "Epoch 624/1000\n",
      "102/102 - 0s - loss: 0.1671 - accuracy: 0.9456\n",
      "Epoch 625/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9459\n",
      "Epoch 626/1000\n",
      "102/102 - 0s - loss: 0.1671 - accuracy: 0.9459\n",
      "Epoch 627/1000\n",
      "102/102 - 0s - loss: 0.1671 - accuracy: 0.9459\n",
      "Epoch 628/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9456\n",
      "Epoch 629/1000\n",
      "102/102 - 0s - loss: 0.1672 - accuracy: 0.9459\n",
      "Epoch 630/1000\n",
      "102/102 - 0s - loss: 0.1671 - accuracy: 0.9462\n",
      "Epoch 631/1000\n",
      "102/102 - 0s - loss: 0.1672 - accuracy: 0.9459\n",
      "Epoch 632/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9462\n",
      "Epoch 633/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9456\n",
      "Epoch 634/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9453\n",
      "Epoch 635/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9462\n",
      "Epoch 636/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9462\n",
      "Epoch 637/1000\n",
      "102/102 - 0s - loss: 0.1671 - accuracy: 0.9459\n",
      "Epoch 638/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9459\n",
      "Epoch 639/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9456\n",
      "Epoch 640/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9459\n",
      "Epoch 641/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9465\n",
      "Epoch 642/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9456\n",
      "Epoch 643/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9462\n",
      "Epoch 644/1000\n",
      "102/102 - 0s - loss: 0.1671 - accuracy: 0.9459\n",
      "Epoch 645/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9462\n",
      "Epoch 646/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9462\n",
      "Epoch 647/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9462\n",
      "Epoch 648/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9453\n",
      "Epoch 649/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9459\n",
      "Epoch 650/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9462\n",
      "Epoch 651/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9453\n",
      "Epoch 652/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9453\n",
      "Epoch 653/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9456\n",
      "Epoch 654/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9459\n",
      "Epoch 655/1000\n",
      "102/102 - 0s - loss: 0.1671 - accuracy: 0.9459\n",
      "Epoch 656/1000\n",
      "102/102 - 0s - loss: 0.1671 - accuracy: 0.9456\n",
      "Epoch 657/1000\n",
      "102/102 - 0s - loss: 0.1671 - accuracy: 0.9456\n",
      "Epoch 658/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9456\n",
      "Epoch 659/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9462\n",
      "Epoch 660/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9459\n",
      "Epoch 661/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9456\n",
      "Epoch 662/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9462\n",
      "Epoch 663/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9453\n",
      "Epoch 664/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9456\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 665/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9453\n",
      "Epoch 666/1000\n",
      "102/102 - 0s - loss: 0.1671 - accuracy: 0.9459\n",
      "Epoch 667/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9462\n",
      "Epoch 668/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9456\n",
      "Epoch 669/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9459\n",
      "Epoch 670/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9459\n",
      "Epoch 671/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9462\n",
      "Epoch 672/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9453\n",
      "Epoch 673/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9459\n",
      "Epoch 674/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9459\n",
      "Epoch 675/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9459\n",
      "Epoch 676/1000\n",
      "102/102 - 0s - loss: 0.1671 - accuracy: 0.9459\n",
      "Epoch 677/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9462\n",
      "Epoch 678/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9462\n",
      "Epoch 679/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9459\n",
      "Epoch 680/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9462\n",
      "Epoch 681/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9462\n",
      "Epoch 682/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9459\n",
      "Epoch 683/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9459\n",
      "Epoch 684/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9453\n",
      "Epoch 685/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9459\n",
      "Epoch 686/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9453\n",
      "Epoch 687/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9462\n",
      "Epoch 688/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9453\n",
      "Epoch 689/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9453\n",
      "Epoch 690/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9456\n",
      "Epoch 691/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9459\n",
      "Epoch 692/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9459\n",
      "Epoch 693/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9459\n",
      "Epoch 694/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9459\n",
      "Epoch 695/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9459\n",
      "Epoch 696/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9459\n",
      "Epoch 697/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9459\n",
      "Epoch 698/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9462\n",
      "Epoch 699/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9459\n",
      "Epoch 700/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9462\n",
      "Epoch 701/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9462\n",
      "Epoch 702/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9462\n",
      "Epoch 703/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9462\n",
      "Epoch 704/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9459\n",
      "Epoch 705/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9462\n",
      "Epoch 706/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9459\n",
      "Epoch 707/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9462\n",
      "Epoch 708/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9459\n",
      "Epoch 709/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9462\n",
      "Epoch 710/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9456\n",
      "Epoch 711/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9462\n",
      "Epoch 712/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9456\n",
      "Epoch 713/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9459\n",
      "Epoch 714/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9459\n",
      "Epoch 715/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9459\n",
      "Epoch 716/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9459\n",
      "Epoch 717/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9459\n",
      "Epoch 718/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9459\n",
      "Epoch 719/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9456\n",
      "Epoch 720/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9462\n",
      "Epoch 721/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9462\n",
      "Epoch 722/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9459\n",
      "Epoch 723/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9459\n",
      "Epoch 724/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9456\n",
      "Epoch 725/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9459\n",
      "Epoch 726/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9462\n",
      "Epoch 727/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9459\n",
      "Epoch 728/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9459\n",
      "Epoch 729/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9459\n",
      "Epoch 730/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9459\n",
      "Epoch 731/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9459\n",
      "Epoch 732/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9462\n",
      "Epoch 733/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9462\n",
      "Epoch 734/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9456\n",
      "Epoch 735/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9462\n",
      "Epoch 736/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9459\n",
      "Epoch 737/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9462\n",
      "Epoch 738/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9462\n",
      "Epoch 739/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9459\n",
      "Epoch 740/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9459\n",
      "Epoch 741/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9459\n",
      "Epoch 742/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9462\n",
      "Epoch 743/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9456\n",
      "Epoch 744/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9459\n",
      "Epoch 745/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9459\n",
      "Epoch 746/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9459\n",
      "Epoch 747/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9462\n",
      "Epoch 748/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9462\n",
      "Epoch 749/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9459\n",
      "Epoch 750/1000\n",
      "102/102 - 0s - loss: 0.1666 - accuracy: 0.9459\n",
      "Epoch 751/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9462\n",
      "Epoch 752/1000\n",
      "102/102 - 0s - loss: 0.1672 - accuracy: 0.9456\n",
      "Epoch 753/1000\n",
      "102/102 - 0s - loss: 0.1670 - accuracy: 0.9459\n",
      "Epoch 754/1000\n",
      "102/102 - 0s - loss: 0.1669 - accuracy: 0.9462\n",
      "Epoch 755/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9459\n",
      "Epoch 756/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9462\n",
      "Epoch 757/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9459\n",
      "Epoch 758/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9459\n",
      "Epoch 759/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9459\n",
      "Epoch 760/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9462\n",
      "Epoch 761/1000\n",
      "102/102 - 0s - loss: 0.1666 - accuracy: 0.9462\n",
      "Epoch 762/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9456\n",
      "Epoch 763/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9462\n",
      "Epoch 764/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9462\n",
      "Epoch 765/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9462\n",
      "Epoch 766/1000\n",
      "102/102 - 0s - loss: 0.1668 - accuracy: 0.9462\n",
      "Epoch 767/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9459\n",
      "Epoch 768/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9462\n",
      "Epoch 769/1000\n",
      "102/102 - 0s - loss: 0.1666 - accuracy: 0.9459\n",
      "Epoch 770/1000\n",
      "102/102 - 0s - loss: 0.1666 - accuracy: 0.9459\n",
      "Epoch 771/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9456\n",
      "Epoch 772/1000\n",
      "102/102 - 0s - loss: 0.1666 - accuracy: 0.9459\n",
      "Epoch 773/1000\n",
      "102/102 - 0s - loss: 0.1666 - accuracy: 0.9459\n",
      "Epoch 774/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9459\n",
      "Epoch 775/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9459\n",
      "Epoch 776/1000\n",
      "102/102 - 0s - loss: 0.1666 - accuracy: 0.9459\n",
      "Epoch 777/1000\n",
      "102/102 - 0s - loss: 0.1666 - accuracy: 0.9459\n",
      "Epoch 778/1000\n",
      "102/102 - 0s - loss: 0.1665 - accuracy: 0.9459\n",
      "Epoch 779/1000\n",
      "102/102 - 0s - loss: 0.1665 - accuracy: 0.9462\n",
      "Epoch 780/1000\n",
      "102/102 - 0s - loss: 0.1665 - accuracy: 0.9459\n",
      "Epoch 781/1000\n",
      "102/102 - 0s - loss: 0.1665 - accuracy: 0.9456\n",
      "Epoch 782/1000\n",
      "102/102 - 0s - loss: 0.1665 - accuracy: 0.9456\n",
      "Epoch 783/1000\n",
      "102/102 - 0s - loss: 0.1665 - accuracy: 0.9456\n",
      "Epoch 784/1000\n",
      "102/102 - 0s - loss: 0.1665 - accuracy: 0.9459\n",
      "Epoch 785/1000\n",
      "102/102 - 0s - loss: 0.1664 - accuracy: 0.9456\n",
      "Epoch 786/1000\n",
      "102/102 - 0s - loss: 0.1665 - accuracy: 0.9456\n",
      "Epoch 787/1000\n",
      "102/102 - 0s - loss: 0.1666 - accuracy: 0.9459\n",
      "Epoch 788/1000\n",
      "102/102 - 0s - loss: 0.1667 - accuracy: 0.9453\n",
      "Epoch 789/1000\n",
      "102/102 - 0s - loss: 0.1665 - accuracy: 0.9456\n",
      "Epoch 790/1000\n",
      "102/102 - 0s - loss: 0.1665 - accuracy: 0.9456\n",
      "Epoch 791/1000\n",
      "102/102 - 0s - loss: 0.1665 - accuracy: 0.9456\n",
      "Epoch 792/1000\n",
      "102/102 - 0s - loss: 0.1666 - accuracy: 0.9453\n",
      "Epoch 793/1000\n",
      "102/102 - 0s - loss: 0.1665 - accuracy: 0.9456\n",
      "Epoch 794/1000\n",
      "102/102 - 0s - loss: 0.1664 - accuracy: 0.9456\n",
      "Epoch 795/1000\n",
      "102/102 - 0s - loss: 0.1665 - accuracy: 0.9456\n",
      "Epoch 796/1000\n",
      "102/102 - 0s - loss: 0.1664 - accuracy: 0.9456\n",
      "Epoch 797/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "102/102 - 0s - loss: 0.1665 - accuracy: 0.9453\n",
      "Epoch 798/1000\n",
      "102/102 - 0s - loss: 0.1664 - accuracy: 0.9456\n",
      "Epoch 799/1000\n",
      "102/102 - 0s - loss: 0.1665 - accuracy: 0.9453\n",
      "Epoch 800/1000\n",
      "102/102 - 0s - loss: 0.1664 - accuracy: 0.9456\n",
      "Epoch 801/1000\n",
      "102/102 - 0s - loss: 0.1665 - accuracy: 0.9456\n",
      "Epoch 802/1000\n",
      "102/102 - 0s - loss: 0.1663 - accuracy: 0.9456\n",
      "Epoch 803/1000\n",
      "102/102 - 0s - loss: 0.1664 - accuracy: 0.9459\n",
      "Epoch 804/1000\n",
      "102/102 - 0s - loss: 0.1665 - accuracy: 0.9456\n",
      "Epoch 805/1000\n",
      "102/102 - 0s - loss: 0.1664 - accuracy: 0.9456\n",
      "Epoch 806/1000\n",
      "102/102 - 0s - loss: 0.1664 - accuracy: 0.9459\n",
      "Epoch 807/1000\n",
      "102/102 - 0s - loss: 0.1664 - accuracy: 0.9453\n",
      "Epoch 808/1000\n",
      "102/102 - 0s - loss: 0.1663 - accuracy: 0.9459\n",
      "Epoch 809/1000\n",
      "102/102 - 0s - loss: 0.1663 - accuracy: 0.9459\n",
      "Epoch 810/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 811/1000\n",
      "102/102 - 0s - loss: 0.1663 - accuracy: 0.9459\n",
      "Epoch 812/1000\n",
      "102/102 - 0s - loss: 0.1663 - accuracy: 0.9459\n",
      "Epoch 813/1000\n",
      "102/102 - 0s - loss: 0.1663 - accuracy: 0.9459\n",
      "Epoch 814/1000\n",
      "102/102 - 0s - loss: 0.1664 - accuracy: 0.9459\n",
      "Epoch 815/1000\n",
      "102/102 - 0s - loss: 0.1663 - accuracy: 0.9459\n",
      "Epoch 816/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 817/1000\n",
      "102/102 - 0s - loss: 0.1664 - accuracy: 0.9459\n",
      "Epoch 818/1000\n",
      "102/102 - 0s - loss: 0.1663 - accuracy: 0.9459\n",
      "Epoch 819/1000\n",
      "102/102 - 0s - loss: 0.1663 - accuracy: 0.9459\n",
      "Epoch 820/1000\n",
      "102/102 - 0s - loss: 0.1663 - accuracy: 0.9456\n",
      "Epoch 821/1000\n",
      "102/102 - 0s - loss: 0.1664 - accuracy: 0.9459\n",
      "Epoch 822/1000\n",
      "102/102 - 0s - loss: 0.1663 - accuracy: 0.9459\n",
      "Epoch 823/1000\n",
      "102/102 - 0s - loss: 0.1663 - accuracy: 0.9459\n",
      "Epoch 824/1000\n",
      "102/102 - 0s - loss: 0.1664 - accuracy: 0.9459\n",
      "Epoch 825/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 826/1000\n",
      "102/102 - 0s - loss: 0.1664 - accuracy: 0.9453\n",
      "Epoch 827/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 828/1000\n",
      "102/102 - 0s - loss: 0.1663 - accuracy: 0.9459\n",
      "Epoch 829/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 830/1000\n",
      "102/102 - 0s - loss: 0.1663 - accuracy: 0.9459\n",
      "Epoch 831/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9456\n",
      "Epoch 832/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 833/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 834/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 835/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 836/1000\n",
      "102/102 - 0s - loss: 0.1663 - accuracy: 0.9456\n",
      "Epoch 837/1000\n",
      "102/102 - 0s - loss: 0.1663 - accuracy: 0.9456\n",
      "Epoch 838/1000\n",
      "102/102 - 0s - loss: 0.1663 - accuracy: 0.9459\n",
      "Epoch 839/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 840/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 841/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 842/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 843/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 844/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 845/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 846/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 847/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9456\n",
      "Epoch 848/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 849/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 850/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9456\n",
      "Epoch 851/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 852/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 853/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 854/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9456\n",
      "Epoch 855/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 856/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 857/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 858/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 859/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 860/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 861/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 862/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 863/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 864/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 865/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 866/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 867/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 868/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 869/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 870/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 871/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 872/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 873/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 874/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 875/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 876/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 877/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 878/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 879/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 880/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 881/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 882/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 883/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 884/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 885/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 886/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 887/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 888/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 889/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 890/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 891/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 892/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 893/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 894/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 895/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 896/1000\n",
      "102/102 - 0s - loss: 0.1663 - accuracy: 0.9462\n",
      "Epoch 897/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 898/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 899/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9462\n",
      "Epoch 900/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 901/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 902/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 903/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 904/1000\n",
      "102/102 - 0s - loss: 0.1663 - accuracy: 0.9462\n",
      "Epoch 905/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 906/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 907/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 908/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 909/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 910/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 911/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 912/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 913/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 914/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 915/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 916/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 917/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 918/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 919/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 920/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 921/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 922/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 923/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 924/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 925/1000\n",
      "102/102 - 0s - loss: 0.1666 - accuracy: 0.9453\n",
      "Epoch 926/1000\n",
      "102/102 - 0s - loss: 0.1663 - accuracy: 0.9465\n",
      "Epoch 927/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 928/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 929/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 930/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 931/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 932/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 933/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 934/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 935/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 936/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 937/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 938/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 939/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 940/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 941/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 942/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 943/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 944/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9462\n",
      "Epoch 945/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 946/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 947/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 948/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9462\n",
      "Epoch 949/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 950/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 951/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 952/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 953/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 954/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 955/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 956/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 957/1000\n",
      "102/102 - 0s - loss: 0.1659 - accuracy: 0.9459\n",
      "Epoch 958/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 959/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 960/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9462\n",
      "Epoch 961/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 962/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 963/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 964/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 965/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 966/1000\n",
      "102/102 - 0s - loss: 0.1659 - accuracy: 0.9459\n",
      "Epoch 967/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 968/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 969/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 970/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 971/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 972/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 973/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 974/1000\n",
      "102/102 - 0s - loss: 0.1659 - accuracy: 0.9459\n",
      "Epoch 975/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 976/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 977/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9462\n",
      "Epoch 978/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 979/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 980/1000\n",
      "102/102 - 0s - loss: 0.1662 - accuracy: 0.9459\n",
      "Epoch 981/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 982/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 983/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 984/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 985/1000\n",
      "102/102 - 0s - loss: 0.1659 - accuracy: 0.9459\n",
      "Epoch 986/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 987/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9462\n",
      "Epoch 988/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 989/1000\n",
      "102/102 - 0s - loss: 0.1659 - accuracy: 0.9459\n",
      "Epoch 990/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 991/1000\n",
      "102/102 - 0s - loss: 0.1661 - accuracy: 0.9459\n",
      "Epoch 992/1000\n",
      "102/102 - 0s - loss: 0.1659 - accuracy: 0.9459\n",
      "Epoch 993/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 994/1000\n",
      "102/102 - 0s - loss: 0.1659 - accuracy: 0.9459\n",
      "Epoch 995/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 996/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 997/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 998/1000\n",
      "102/102 - 0s - loss: 0.1659 - accuracy: 0.9459\n",
      "Epoch 999/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n",
      "Epoch 1000/1000\n",
      "102/102 - 0s - loss: 0.1660 - accuracy: 0.9459\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.keras.callbacks.History at 0x17c41a31f70>"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Fit (train) the model\n",
    "model.fit(\n",
    "    X_train_scaled,\n",
    "    y_train_categorical,\n",
    "    epochs=1000,\n",
    "    shuffle=True,\n",
    "    verbose=2\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Quantifying the Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "34/34 - 0s - loss: 0.2273 - accuracy: 0.9333\n",
      "Loss: 0.2273392379283905, Accuracy: 0.9332715272903442\n"
     ]
    }
   ],
   "source": [
    "# Evaluate the model using the testing data\n",
    "model_loss, model_accuracy = model.evaluate(\n",
    "    X_test_scaled, y_test_categorical, verbose=2)\n",
    "print(f\"Loss: {model_loss}, Accuracy: {model_accuracy}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}