Credit Card Fraud.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Credit Card Fraud Detection"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Context\n",
    "As presented by the 5th 1056Lab Data Analytics Competition\n",
    "Fraud detection in financial transactions is one of the most important problems in financial companies.\n",
    "\n",
    "The original dataset is in Kaggle Datasets.\n",
    "This data is about fraud detection in credit card transactions. The data was made by credit cards in September 2013 by European cardholders. The dataset is highly unbalanced, the positive class (frauds) account for 0.172% of all transactions.\n",
    "It contains only numerical input variables which are the result of a PCA transformation. The original features and more background information about the data is not provided.\n",
    "The dataset contains 284,807 instances, 492 instances are fraudulent, the remaining 284,315 instances are genuine.\n",
    "Training Dat and Test Data\n",
    "The training data set is the former 198,365 (70%) instances, 383 represents fraud transactions, 197,982 transactions are genuine.\n",
    "The test data is the later 86,442 (30%) transactions."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## My Approach is quite straitforward\n",
    "\n",
    "The main issue with this task is the unbalanced data, so i will present two approaches, one with a balanced dataset and the other with the original (unbalanced) dataset.\n",
    "1. I will do a bit of exploratory data analysis\n",
    "2. Then do some feature engineering\n",
    "3. Then Balance the dataset\n",
    "3. Apply some ML models with default parameters; models like logistic regression, XGboost, SVC, KNN etc.\n",
    "4. I will examine the accuracy\n",
    "5. Apply deep learning model, validate then compare the model results\n",
    "6. Then Make some prediction on the test\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Import the relevant libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import preprocessing\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.base import BaseEstimator, TransformerMixin\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.metrics import accuracy_score, confusion_matrix\n",
    "from sklearn import svm, tree\n",
    "import xgboost\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.ensemble import RandomForestClassifier, AdaBoostClassifier\n",
    "from sklearn.neural_network import MLPClassifier\n",
    "from sklearn import metrics\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "import pickle\n",
    "%matplotlib inline\n",
    "import datetime \n",
    "import time \n",
    "import tensorflow as tf\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load the data\n",
    "data  = pd.read_csv('train.csv')\n",
    "test  = pd.read_csv('test.csv')\n",
    "submission = test[['ID']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 198365 entries, 0 to 198364\n",
      "Data columns (total 32 columns):\n",
      " #   Column  Non-Null Count   Dtype  \n",
      "---  ------  --------------   -----  \n",
      " 0   ID      198365 non-null  int64  \n",
      " 1   Time    198365 non-null  float64\n",
      " 2   V1      198365 non-null  float64\n",
      " 3   V2      198365 non-null  float64\n",
      " 4   V3      198365 non-null  float64\n",
      " 5   V4      198365 non-null  float64\n",
      " 6   V5      198365 non-null  float64\n",
      " 7   V6      198365 non-null  float64\n",
      " 8   V7      198365 non-null  float64\n",
      " 9   V8      198365 non-null  float64\n",
      " 10  V9      198365 non-null  float64\n",
      " 11  V10     198365 non-null  float64\n",
      " 12  V11     198365 non-null  float64\n",
      " 13  V12     198365 non-null  float64\n",
      " 14  V13     198365 non-null  float64\n",
      " 15  V14     198365 non-null  float64\n",
      " 16  V15     198365 non-null  float64\n",
      " 17  V16     198365 non-null  float64\n",
      " 18  V17     198365 non-null  float64\n",
      " 19  V18     198365 non-null  float64\n",
      " 20  V19     198365 non-null  float64\n",
      " 21  V20     198365 non-null  float64\n",
      " 22  V21     198365 non-null  float64\n",
      " 23  V22     198365 non-null  float64\n",
      " 24  V23     198365 non-null  float64\n",
      " 25  V24     198365 non-null  float64\n",
      " 26  V25     198365 non-null  float64\n",
      " 27  V26     198365 non-null  float64\n",
      " 28  V27     198365 non-null  float64\n",
      " 29  V28     198365 non-null  float64\n",
      " 30  Amount  198365 non-null  float64\n",
      " 31  Class   198365 non-null  int64  \n",
      "dtypes: float64(30), int64(2)\n",
      "memory usage: 48.4 MB\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 86442 entries, 0 to 86441\n",
      "Data columns (total 31 columns):\n",
      " #   Column  Non-Null Count  Dtype  \n",
      "---  ------  --------------  -----  \n",
      " 0   ID      86442 non-null  int64  \n",
      " 1   Time    86442 non-null  int64  \n",
      " 2   V1      86442 non-null  float64\n",
      " 3   V2      86442 non-null  float64\n",
      " 4   V3      86442 non-null  float64\n",
      " 5   V4      86442 non-null  float64\n",
      " 6   V5      86442 non-null  float64\n",
      " 7   V6      86442 non-null  float64\n",
      " 8   V7      86442 non-null  float64\n",
      " 9   V8      86442 non-null  float64\n",
      " 10  V9      86442 non-null  float64\n",
      " 11  V10     86442 non-null  float64\n",
      " 12  V11     86442 non-null  float64\n",
      " 13  V12     86442 non-null  float64\n",
      " 14  V13     86442 non-null  float64\n",
      " 15  V14     86442 non-null  float64\n",
      " 16  V15     86442 non-null  float64\n",
      " 17  V16     86442 non-null  float64\n",
      " 18  V17     86442 non-null  float64\n",
      " 19  V18     86442 non-null  float64\n",
      " 20  V19     86442 non-null  float64\n",
      " 21  V20     86442 non-null  float64\n",
      " 22  V21     86442 non-null  float64\n",
      " 23  V22     86442 non-null  float64\n",
      " 24  V23     86442 non-null  float64\n",
      " 25  V24     86442 non-null  float64\n",
      " 26  V25     86442 non-null  float64\n",
      " 27  V26     86442 non-null  float64\n",
      " 28  V27     86442 non-null  float64\n",
      " 29  V28     86442 non-null  float64\n",
      " 30  Amount  86442 non-null  float64\n",
      "dtypes: float64(29), int64(2)\n",
      "memory usage: 20.4 MB\n"
     ]
    }
   ],
   "source": [
    "#display some information about data\n",
    "data.info()\n",
    "test.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ID        0\n",
       "Time      0\n",
       "V1        0\n",
       "V2        0\n",
       "V3        0\n",
       "V4        0\n",
       "V5        0\n",
       "V6        0\n",
       "V7        0\n",
       "V8        0\n",
       "V9        0\n",
       "V10       0\n",
       "V11       0\n",
       "V12       0\n",
       "V13       0\n",
       "V14       0\n",
       "V15       0\n",
       "V16       0\n",
       "V17       0\n",
       "V18       0\n",
       "V19       0\n",
       "V20       0\n",
       "V21       0\n",
       "V22       0\n",
       "V23       0\n",
       "V24       0\n",
       "V25       0\n",
       "V26       0\n",
       "V27       0\n",
       "V28       0\n",
       "Amount    0\n",
       "Class     0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# data.isnull() # shows a df with the information whether a data point is null \n",
    "# Since True = the data point is missing, while False = the data point is not missing, we can sum them\n",
    "# This will give us the total number of missing values feature-wise\n",
    "data.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1440 with 36 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# As we can see there are no missing data\n",
    "\n",
    "# so let take a look at the spread of each column\n",
    "# Visulazing the distibution of the data for every feature\n",
    "data.hist(linewidth=1, histtype='stepfilled', facecolor='g', figsize=(20, 20));"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "scrolled": false
   },
   "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>Time</th>\n",
       "      <th>V1</th>\n",
       "      <th>V2</th>\n",
       "      <th>V3</th>\n",
       "      <th>V4</th>\n",
       "      <th>V5</th>\n",
       "      <th>V6</th>\n",
       "      <th>V7</th>\n",
       "      <th>V8</th>\n",
       "      <th>...</th>\n",
       "      <th>V21</th>\n",
       "      <th>V22</th>\n",
       "      <th>V23</th>\n",
       "      <th>V24</th>\n",
       "      <th>V25</th>\n",
       "      <th>V26</th>\n",
       "      <th>V27</th>\n",
       "      <th>V28</th>\n",
       "      <th>Amount</th>\n",
       "      <th>Class</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>198365.000000</td>\n",
       "      <td>198365.000000</td>\n",
       "      <td>198365.000000</td>\n",
       "      <td>198365.000000</td>\n",
       "      <td>198365.000000</td>\n",
       "      <td>198365.000000</td>\n",
       "      <td>198365.000000</td>\n",
       "      <td>198365.000000</td>\n",
       "      <td>198365.000000</td>\n",
       "      <td>198365.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>198365.000000</td>\n",
       "      <td>198365.000000</td>\n",
       "      <td>198365.000000</td>\n",
       "      <td>198365.000000</td>\n",
       "      <td>198365.000000</td>\n",
       "      <td>198365.000000</td>\n",
       "      <td>198365.000000</td>\n",
       "      <td>198365.000000</td>\n",
       "      <td>198365.000000</td>\n",
       "      <td>198365.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>99183.000000</td>\n",
       "      <td>70130.934414</td>\n",
       "      <td>-0.107339</td>\n",
       "      <td>0.004431</td>\n",
       "      <td>0.325762</td>\n",
       "      <td>0.077514</td>\n",
       "      <td>-0.113668</td>\n",
       "      <td>0.046607</td>\n",
       "      <td>-0.051740</td>\n",
       "      <td>0.017719</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.014966</td>\n",
       "      <td>-0.049185</td>\n",
       "      <td>-0.015766</td>\n",
       "      <td>0.004333</td>\n",
       "      <td>0.061965</td>\n",
       "      <td>0.006424</td>\n",
       "      <td>0.001587</td>\n",
       "      <td>0.001976</td>\n",
       "      <td>89.703525</td>\n",
       "      <td>0.001931</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>57263.187411</td>\n",
       "      <td>34305.103318</td>\n",
       "      <td>1.891453</td>\n",
       "      <td>1.622666</td>\n",
       "      <td>1.446295</td>\n",
       "      <td>1.398805</td>\n",
       "      <td>1.360659</td>\n",
       "      <td>1.310315</td>\n",
       "      <td>1.216025</td>\n",
       "      <td>1.210508</td>\n",
       "      <td>...</td>\n",
       "      <td>0.738678</td>\n",
       "      <td>0.690646</td>\n",
       "      <td>0.610795</td>\n",
       "      <td>0.603471</td>\n",
       "      <td>0.490441</td>\n",
       "      <td>0.488446</td>\n",
       "      <td>0.394026</td>\n",
       "      <td>0.314154</td>\n",
       "      <td>248.680051</td>\n",
       "      <td>0.043898</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-56.407510</td>\n",
       "      <td>-72.715728</td>\n",
       "      <td>-33.680984</td>\n",
       "      <td>-5.683171</td>\n",
       "      <td>-42.147898</td>\n",
       "      <td>-26.160506</td>\n",
       "      <td>-43.557242</td>\n",
       "      <td>-73.216718</td>\n",
       "      <td>...</td>\n",
       "      <td>-34.830382</td>\n",
       "      <td>-10.933144</td>\n",
       "      <td>-44.807735</td>\n",
       "      <td>-2.836627</td>\n",
       "      <td>-10.295397</td>\n",
       "      <td>-2.604551</td>\n",
       "      <td>-22.565679</td>\n",
       "      <td>-11.710896</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>49592.000000</td>\n",
       "      <td>44128.000000</td>\n",
       "      <td>-0.962685</td>\n",
       "      <td>-0.580147</td>\n",
       "      <td>-0.327270</td>\n",
       "      <td>-0.790975</td>\n",
       "      <td>-0.785452</td>\n",
       "      <td>-0.709602</td>\n",
       "      <td>-0.576733</td>\n",
       "      <td>-0.180360</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.227188</td>\n",
       "      <td>-0.538206</td>\n",
       "      <td>-0.168622</td>\n",
       "      <td>-0.341607</td>\n",
       "      <td>-0.238794</td>\n",
       "      <td>-0.331187</td>\n",
       "      <td>-0.067367</td>\n",
       "      <td>-0.041312</td>\n",
       "      <td>5.990000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>99183.000000</td>\n",
       "      <td>67005.000000</td>\n",
       "      <td>-0.094105</td>\n",
       "      <td>0.082136</td>\n",
       "      <td>0.496715</td>\n",
       "      <td>0.078139</td>\n",
       "      <td>-0.171469</td>\n",
       "      <td>-0.218506</td>\n",
       "      <td>-0.010989</td>\n",
       "      <td>0.043647</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.043309</td>\n",
       "      <td>-0.035961</td>\n",
       "      <td>-0.029126</td>\n",
       "      <td>0.052775</td>\n",
       "      <td>0.104492</td>\n",
       "      <td>-0.061019</td>\n",
       "      <td>0.006036</td>\n",
       "      <td>0.018528</td>\n",
       "      <td>22.990000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>148774.000000</td>\n",
       "      <td>90198.000000</td>\n",
       "      <td>1.213058</td>\n",
       "      <td>0.783763</td>\n",
       "      <td>1.225320</td>\n",
       "      <td>0.889229</td>\n",
       "      <td>0.462401</td>\n",
       "      <td>0.444661</td>\n",
       "      <td>0.499504</td>\n",
       "      <td>0.339098</td>\n",
       "      <td>...</td>\n",
       "      <td>0.150027</td>\n",
       "      <td>0.421133</td>\n",
       "      <td>0.114667</td>\n",
       "      <td>0.421534</td>\n",
       "      <td>0.384995</td>\n",
       "      <td>0.261761</td>\n",
       "      <td>0.089128</td>\n",
       "      <td>0.078235</td>\n",
       "      <td>79.200000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>198365.000000</td>\n",
       "      <td>132458.000000</td>\n",
       "      <td>2.454930</td>\n",
       "      <td>22.057729</td>\n",
       "      <td>9.382558</td>\n",
       "      <td>16.875344</td>\n",
       "      <td>34.801666</td>\n",
       "      <td>22.529298</td>\n",
       "      <td>36.677268</td>\n",
       "      <td>20.007208</td>\n",
       "      <td>...</td>\n",
       "      <td>27.202839</td>\n",
       "      <td>10.503090</td>\n",
       "      <td>19.002942</td>\n",
       "      <td>4.022866</td>\n",
       "      <td>7.519589</td>\n",
       "      <td>3.517346</td>\n",
       "      <td>12.152401</td>\n",
       "      <td>33.847808</td>\n",
       "      <td>19656.530000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 32 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                  ID           Time             V1             V2  \\\n",
       "count  198365.000000  198365.000000  198365.000000  198365.000000   \n",
       "mean    99183.000000   70130.934414      -0.107339       0.004431   \n",
       "std     57263.187411   34305.103318       1.891453       1.622666   \n",
       "min         1.000000       0.000000     -56.407510     -72.715728   \n",
       "25%     49592.000000   44128.000000      -0.962685      -0.580147   \n",
       "50%     99183.000000   67005.000000      -0.094105       0.082136   \n",
       "75%    148774.000000   90198.000000       1.213058       0.783763   \n",
       "max    198365.000000  132458.000000       2.454930      22.057729   \n",
       "\n",
       "                  V3             V4             V5             V6  \\\n",
       "count  198365.000000  198365.000000  198365.000000  198365.000000   \n",
       "mean        0.325762       0.077514      -0.113668       0.046607   \n",
       "std         1.446295       1.398805       1.360659       1.310315   \n",
       "min       -33.680984      -5.683171     -42.147898     -26.160506   \n",
       "25%        -0.327270      -0.790975      -0.785452      -0.709602   \n",
       "50%         0.496715       0.078139      -0.171469      -0.218506   \n",
       "75%         1.225320       0.889229       0.462401       0.444661   \n",
       "max         9.382558      16.875344      34.801666      22.529298   \n",
       "\n",
       "                  V7             V8  ...            V21            V22  \\\n",
       "count  198365.000000  198365.000000  ...  198365.000000  198365.000000   \n",
       "mean       -0.051740       0.017719  ...      -0.014966      -0.049185   \n",
       "std         1.216025       1.210508  ...       0.738678       0.690646   \n",
       "min       -43.557242     -73.216718  ...     -34.830382     -10.933144   \n",
       "25%        -0.576733      -0.180360  ...      -0.227188      -0.538206   \n",
       "50%        -0.010989       0.043647  ...      -0.043309      -0.035961   \n",
       "75%         0.499504       0.339098  ...       0.150027       0.421133   \n",
       "max        36.677268      20.007208  ...      27.202839      10.503090   \n",
       "\n",
       "                 V23            V24            V25            V26  \\\n",
       "count  198365.000000  198365.000000  198365.000000  198365.000000   \n",
       "mean       -0.015766       0.004333       0.061965       0.006424   \n",
       "std         0.610795       0.603471       0.490441       0.488446   \n",
       "min       -44.807735      -2.836627     -10.295397      -2.604551   \n",
       "25%        -0.168622      -0.341607      -0.238794      -0.331187   \n",
       "50%        -0.029126       0.052775       0.104492      -0.061019   \n",
       "75%         0.114667       0.421534       0.384995       0.261761   \n",
       "max        19.002942       4.022866       7.519589       3.517346   \n",
       "\n",
       "                 V27            V28         Amount          Class  \n",
       "count  198365.000000  198365.000000  198365.000000  198365.000000  \n",
       "mean        0.001587       0.001976      89.703525       0.001931  \n",
       "std         0.394026       0.314154     248.680051       0.043898  \n",
       "min       -22.565679     -11.710896       0.000000       0.000000  \n",
       "25%        -0.067367      -0.041312       5.990000       0.000000  \n",
       "50%         0.006036       0.018528      22.990000       0.000000  \n",
       "75%         0.089128       0.078235      79.200000       0.000000  \n",
       "max        12.152401      33.847808   19656.530000       1.000000  \n",
       "\n",
       "[8 rows x 32 columns]"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#lets look at the basic description of the data\n",
    "data.describe()\n",
    "#test.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Feature Engineering"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Convert Time column into bank of hours in a day"
   ]
  },
  {
   "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>ID</th>\n",
       "      <th>Time</th>\n",
       "      <th>V1</th>\n",
       "      <th>V2</th>\n",
       "      <th>V3</th>\n",
       "      <th>V4</th>\n",
       "      <th>V5</th>\n",
       "      <th>V6</th>\n",
       "      <th>V7</th>\n",
       "      <th>V8</th>\n",
       "      <th>...</th>\n",
       "      <th>V22</th>\n",
       "      <th>V23</th>\n",
       "      <th>V24</th>\n",
       "      <th>V25</th>\n",
       "      <th>V26</th>\n",
       "      <th>V27</th>\n",
       "      <th>V28</th>\n",
       "      <th>Amount</th>\n",
       "      <th>Class</th>\n",
       "      <th>HourBank</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-1.359807</td>\n",
       "      <td>-0.072781</td>\n",
       "      <td>2.536347</td>\n",
       "      <td>1.378155</td>\n",
       "      <td>-0.338321</td>\n",
       "      <td>0.462388</td>\n",
       "      <td>0.239599</td>\n",
       "      <td>0.098698</td>\n",
       "      <td>...</td>\n",
       "      <td>0.277838</td>\n",
       "      <td>-0.110474</td>\n",
       "      <td>0.066928</td>\n",
       "      <td>0.128539</td>\n",
       "      <td>-0.189115</td>\n",
       "      <td>0.133558</td>\n",
       "      <td>-0.021053</td>\n",
       "      <td>149.62</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.191857</td>\n",
       "      <td>0.266151</td>\n",
       "      <td>0.166480</td>\n",
       "      <td>0.448154</td>\n",
       "      <td>0.060018</td>\n",
       "      <td>-0.082361</td>\n",
       "      <td>-0.078803</td>\n",
       "      <td>0.085102</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.638672</td>\n",
       "      <td>0.101288</td>\n",
       "      <td>-0.339846</td>\n",
       "      <td>0.167170</td>\n",
       "      <td>0.125895</td>\n",
       "      <td>-0.008983</td>\n",
       "      <td>0.014724</td>\n",
       "      <td>2.69</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-1.358354</td>\n",
       "      <td>-1.340163</td>\n",
       "      <td>1.773209</td>\n",
       "      <td>0.379780</td>\n",
       "      <td>-0.503198</td>\n",
       "      <td>1.800499</td>\n",
       "      <td>0.791461</td>\n",
       "      <td>0.247676</td>\n",
       "      <td>...</td>\n",
       "      <td>0.771679</td>\n",
       "      <td>0.909412</td>\n",
       "      <td>-0.689281</td>\n",
       "      <td>-0.327642</td>\n",
       "      <td>-0.139097</td>\n",
       "      <td>-0.055353</td>\n",
       "      <td>-0.059752</td>\n",
       "      <td>378.66</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-0.966272</td>\n",
       "      <td>-0.185226</td>\n",
       "      <td>1.792993</td>\n",
       "      <td>-0.863291</td>\n",
       "      <td>-0.010309</td>\n",
       "      <td>1.247203</td>\n",
       "      <td>0.237609</td>\n",
       "      <td>0.377436</td>\n",
       "      <td>...</td>\n",
       "      <td>0.005274</td>\n",
       "      <td>-0.190321</td>\n",
       "      <td>-1.175575</td>\n",
       "      <td>0.647376</td>\n",
       "      <td>-0.221929</td>\n",
       "      <td>0.062723</td>\n",
       "      <td>0.061458</td>\n",
       "      <td>123.50</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>2.0</td>\n",
       "      <td>-1.158233</td>\n",
       "      <td>0.877737</td>\n",
       "      <td>1.548718</td>\n",
       "      <td>0.403034</td>\n",
       "      <td>-0.407193</td>\n",
       "      <td>0.095921</td>\n",
       "      <td>0.592941</td>\n",
       "      <td>-0.270533</td>\n",
       "      <td>...</td>\n",
       "      <td>0.798278</td>\n",
       "      <td>-0.137458</td>\n",
       "      <td>0.141267</td>\n",
       "      <td>-0.206010</td>\n",
       "      <td>0.502292</td>\n",
       "      <td>0.219422</td>\n",
       "      <td>0.215153</td>\n",
       "      <td>69.99</td>\n",
       "      <td>0</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",
       "      <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>198360</th>\n",
       "      <td>198361</td>\n",
       "      <td>132457.0</td>\n",
       "      <td>-0.474163</td>\n",
       "      <td>1.438815</td>\n",
       "      <td>-0.816746</td>\n",
       "      <td>0.786791</td>\n",
       "      <td>1.067826</td>\n",
       "      <td>-0.207767</td>\n",
       "      <td>1.251817</td>\n",
       "      <td>0.140304</td>\n",
       "      <td>...</td>\n",
       "      <td>0.527323</td>\n",
       "      <td>-0.272449</td>\n",
       "      <td>-1.027952</td>\n",
       "      <td>0.237013</td>\n",
       "      <td>-0.312190</td>\n",
       "      <td>0.304261</td>\n",
       "      <td>0.170698</td>\n",
       "      <td>54.21</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>198361</th>\n",
       "      <td>198362</td>\n",
       "      <td>132457.0</td>\n",
       "      <td>1.501100</td>\n",
       "      <td>-2.329402</td>\n",
       "      <td>-0.028328</td>\n",
       "      <td>-0.579340</td>\n",
       "      <td>-2.269449</td>\n",
       "      <td>-0.208058</td>\n",
       "      <td>-1.307538</td>\n",
       "      <td>0.083778</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.112426</td>\n",
       "      <td>0.103570</td>\n",
       "      <td>-0.060057</td>\n",
       "      <td>-0.973943</td>\n",
       "      <td>1.226613</td>\n",
       "      <td>-0.080066</td>\n",
       "      <td>0.009990</td>\n",
       "      <td>314.90</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>198362</th>\n",
       "      <td>198363</td>\n",
       "      <td>132458.0</td>\n",
       "      <td>1.957960</td>\n",
       "      <td>-0.677669</td>\n",
       "      <td>-0.395253</td>\n",
       "      <td>0.762448</td>\n",
       "      <td>-0.685747</td>\n",
       "      <td>0.201289</td>\n",
       "      <td>-0.854448</td>\n",
       "      <td>0.198542</td>\n",
       "      <td>...</td>\n",
       "      <td>0.907851</td>\n",
       "      <td>0.061666</td>\n",
       "      <td>0.730593</td>\n",
       "      <td>0.001207</td>\n",
       "      <td>0.568559</td>\n",
       "      <td>-0.009303</td>\n",
       "      <td>-0.047854</td>\n",
       "      <td>14.39</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>198363</th>\n",
       "      <td>198364</td>\n",
       "      <td>132458.0</td>\n",
       "      <td>-1.135983</td>\n",
       "      <td>-2.150924</td>\n",
       "      <td>-1.078870</td>\n",
       "      <td>-0.922886</td>\n",
       "      <td>-1.936992</td>\n",
       "      <td>0.819695</td>\n",
       "      <td>3.101553</td>\n",
       "      <td>-0.311864</td>\n",
       "      <td>...</td>\n",
       "      <td>0.629329</td>\n",
       "      <td>1.597087</td>\n",
       "      <td>-0.658563</td>\n",
       "      <td>0.298377</td>\n",
       "      <td>-0.101644</td>\n",
       "      <td>-0.198980</td>\n",
       "      <td>0.084730</td>\n",
       "      <td>804.00</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>198364</th>\n",
       "      <td>198365</td>\n",
       "      <td>132458.0</td>\n",
       "      <td>-0.664842</td>\n",
       "      <td>0.635278</td>\n",
       "      <td>1.398580</td>\n",
       "      <td>-0.412006</td>\n",
       "      <td>0.357441</td>\n",
       "      <td>-0.090934</td>\n",
       "      <td>0.804617</td>\n",
       "      <td>0.076064</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.374581</td>\n",
       "      <td>0.037338</td>\n",
       "      <td>-0.518470</td>\n",
       "      <td>-0.053920</td>\n",
       "      <td>0.334494</td>\n",
       "      <td>-0.036098</td>\n",
       "      <td>0.054529</td>\n",
       "      <td>61.80</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>198365 rows × 33 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            ID      Time        V1        V2        V3        V4        V5  \\\n",
       "0            1       0.0 -1.359807 -0.072781  2.536347  1.378155 -0.338321   \n",
       "1            2       0.0  1.191857  0.266151  0.166480  0.448154  0.060018   \n",
       "2            3       1.0 -1.358354 -1.340163  1.773209  0.379780 -0.503198   \n",
       "3            4       1.0 -0.966272 -0.185226  1.792993 -0.863291 -0.010309   \n",
       "4            5       2.0 -1.158233  0.877737  1.548718  0.403034 -0.407193   \n",
       "...        ...       ...       ...       ...       ...       ...       ...   \n",
       "198360  198361  132457.0 -0.474163  1.438815 -0.816746  0.786791  1.067826   \n",
       "198361  198362  132457.0  1.501100 -2.329402 -0.028328 -0.579340 -2.269449   \n",
       "198362  198363  132458.0  1.957960 -0.677669 -0.395253  0.762448 -0.685747   \n",
       "198363  198364  132458.0 -1.135983 -2.150924 -1.078870 -0.922886 -1.936992   \n",
       "198364  198365  132458.0 -0.664842  0.635278  1.398580 -0.412006  0.357441   \n",
       "\n",
       "              V6        V7        V8  ...       V22       V23       V24  \\\n",
       "0       0.462388  0.239599  0.098698  ...  0.277838 -0.110474  0.066928   \n",
       "1      -0.082361 -0.078803  0.085102  ... -0.638672  0.101288 -0.339846   \n",
       "2       1.800499  0.791461  0.247676  ...  0.771679  0.909412 -0.689281   \n",
       "3       1.247203  0.237609  0.377436  ...  0.005274 -0.190321 -1.175575   \n",
       "4       0.095921  0.592941 -0.270533  ...  0.798278 -0.137458  0.141267   \n",
       "...          ...       ...       ...  ...       ...       ...       ...   \n",
       "198360 -0.207767  1.251817  0.140304  ...  0.527323 -0.272449 -1.027952   \n",
       "198361 -0.208058 -1.307538  0.083778  ... -0.112426  0.103570 -0.060057   \n",
       "198362  0.201289 -0.854448  0.198542  ...  0.907851  0.061666  0.730593   \n",
       "198363  0.819695  3.101553 -0.311864  ...  0.629329  1.597087 -0.658563   \n",
       "198364 -0.090934  0.804617  0.076064  ... -1.374581  0.037338 -0.518470   \n",
       "\n",
       "             V25       V26       V27       V28  Amount  Class  HourBank  \n",
       "0       0.128539 -0.189115  0.133558 -0.021053  149.62      0         0  \n",
       "1       0.167170  0.125895 -0.008983  0.014724    2.69      0         0  \n",
       "2      -0.327642 -0.139097 -0.055353 -0.059752  378.66      0         0  \n",
       "3       0.647376 -0.221929  0.062723  0.061458  123.50      0         0  \n",
       "4      -0.206010  0.502292  0.219422  0.215153   69.99      0         0  \n",
       "...          ...       ...       ...       ...     ...    ...       ...  \n",
       "198360  0.237013 -0.312190  0.304261  0.170698   54.21      0        12  \n",
       "198361 -0.973943  1.226613 -0.080066  0.009990  314.90      0        12  \n",
       "198362  0.001207  0.568559 -0.009303 -0.047854   14.39      0        12  \n",
       "198363  0.298377 -0.101644 -0.198980  0.084730  804.00      0        12  \n",
       "198364 -0.053920  0.334494 -0.036098  0.054529   61.80      0        12  \n",
       "\n",
       "[198365 rows x 33 columns]"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# looking at the datset, we can infer that the datsset contains two days of traction or the record spans a  48hrs period and as \n",
    "# such, I have generated the hour of transaction by assuming that the first seconds of a day is 0 and the last one 86399, Hence:\n",
    "\n",
    "data['HourBank'] = ((np.where(data['Time'] > 86399 , data['Time'] - 86399 , data['Time'])) % (24 *3600) // 3600).astype(int)\n",
    "test['HourBank'] = ((np.where(test['Time'] > 86399 , test['Time'] - 86399 , test['Time'])) % (24 *3600) // 3600).astype(int)   \n",
    "#temptime = np.where(data['Time'] > 86399 , data['Time'] - 86399 , data['Time'])\n",
    "#data['HourBank'] = data['HourBank'].astype(int)\n",
    "#test['HourBank'] = test['HourBank'].astype(int)\n",
    "data\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Checking the distribution of the targets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Class\n",
       "0    197982\n",
       "1       383\n",
       "Name: Class, dtype: int64"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['Class'].groupby(data['Class']).count()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Exploring the distribution of the Class and Amount"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# First we limit the data frame to where a fraudulent activity was identify\n",
    "d = data[data['Class'] == 1]\n",
    "\n",
    "# group by HourBank, then count of fradulent transaction\n",
    "d1 = d[['Class','HourBank']]\n",
    "\n",
    "d1 = d1.groupby(['HourBank']).count()\n",
    "d1.reset_index(level=0, inplace=True)\n",
    "\n",
    "# group by HourBank, then average Amount of fradulent transaction\n",
    "d2 = d[['Amount','HourBank']]\n",
    "# group store and dept by average weekly sales\n",
    "d2 = d2.groupby(['HourBank']).mean()\n",
    "d2.reset_index(level=0, inplace=True)\n",
    "\n",
    "\n",
    "#Lets do a quick plot to visualise the data\n",
    "fig, (ax1, ax2) = plt.subplots(nrows=2, sharex=True, figsize=(10,7))\n",
    "\n",
    "ax1.bar(d1.HourBank, d1.Class)\n",
    "ax1.set_title('Count of Fraudulent transactions by Hour')\n",
    "ax2.bar(d2.HourBank, d2.Amount)\n",
    "ax2.set_title('Average Amount Classed as Fraudulent by Hour')\n",
    "\n",
    "fig.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The plots above shows the distribution of fraudulent transaction over the time(hr) of the day"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "One Hot Encoding of the Hour Banks"
   ]
  },
  {
   "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>V1</th>\n",
       "      <th>V2</th>\n",
       "      <th>V3</th>\n",
       "      <th>V4</th>\n",
       "      <th>V5</th>\n",
       "      <th>V6</th>\n",
       "      <th>V7</th>\n",
       "      <th>V8</th>\n",
       "      <th>V9</th>\n",
       "      <th>V10</th>\n",
       "      <th>...</th>\n",
       "      <th>14</th>\n",
       "      <th>15</th>\n",
       "      <th>16</th>\n",
       "      <th>17</th>\n",
       "      <th>18</th>\n",
       "      <th>19</th>\n",
       "      <th>20</th>\n",
       "      <th>21</th>\n",
       "      <th>22</th>\n",
       "      <th>23</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.359807</td>\n",
       "      <td>-0.072781</td>\n",
       "      <td>2.536347</td>\n",
       "      <td>1.378155</td>\n",
       "      <td>-0.338321</td>\n",
       "      <td>0.462388</td>\n",
       "      <td>0.239599</td>\n",
       "      <td>0.098698</td>\n",
       "      <td>0.363787</td>\n",
       "      <td>0.090794</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.191857</td>\n",
       "      <td>0.266151</td>\n",
       "      <td>0.166480</td>\n",
       "      <td>0.448154</td>\n",
       "      <td>0.060018</td>\n",
       "      <td>-0.082361</td>\n",
       "      <td>-0.078803</td>\n",
       "      <td>0.085102</td>\n",
       "      <td>-0.255425</td>\n",
       "      <td>-0.166974</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-1.358354</td>\n",
       "      <td>-1.340163</td>\n",
       "      <td>1.773209</td>\n",
       "      <td>0.379780</td>\n",
       "      <td>-0.503198</td>\n",
       "      <td>1.800499</td>\n",
       "      <td>0.791461</td>\n",
       "      <td>0.247676</td>\n",
       "      <td>-1.514654</td>\n",
       "      <td>0.207643</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.966272</td>\n",
       "      <td>-0.185226</td>\n",
       "      <td>1.792993</td>\n",
       "      <td>-0.863291</td>\n",
       "      <td>-0.010309</td>\n",
       "      <td>1.247203</td>\n",
       "      <td>0.237609</td>\n",
       "      <td>0.377436</td>\n",
       "      <td>-1.387024</td>\n",
       "      <td>-0.054952</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.158233</td>\n",
       "      <td>0.877737</td>\n",
       "      <td>1.548718</td>\n",
       "      <td>0.403034</td>\n",
       "      <td>-0.407193</td>\n",
       "      <td>0.095921</td>\n",
       "      <td>0.592941</td>\n",
       "      <td>-0.270533</td>\n",
       "      <td>0.817739</td>\n",
       "      <td>0.753074</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</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",
       "      <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>198360</th>\n",
       "      <td>-0.474163</td>\n",
       "      <td>1.438815</td>\n",
       "      <td>-0.816746</td>\n",
       "      <td>0.786791</td>\n",
       "      <td>1.067826</td>\n",
       "      <td>-0.207767</td>\n",
       "      <td>1.251817</td>\n",
       "      <td>0.140304</td>\n",
       "      <td>-0.903642</td>\n",
       "      <td>-0.286039</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>198361</th>\n",
       "      <td>1.501100</td>\n",
       "      <td>-2.329402</td>\n",
       "      <td>-0.028328</td>\n",
       "      <td>-0.579340</td>\n",
       "      <td>-2.269449</td>\n",
       "      <td>-0.208058</td>\n",
       "      <td>-1.307538</td>\n",
       "      <td>0.083778</td>\n",
       "      <td>0.774633</td>\n",
       "      <td>0.584604</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>198362</th>\n",
       "      <td>1.957960</td>\n",
       "      <td>-0.677669</td>\n",
       "      <td>-0.395253</td>\n",
       "      <td>0.762448</td>\n",
       "      <td>-0.685747</td>\n",
       "      <td>0.201289</td>\n",
       "      <td>-0.854448</td>\n",
       "      <td>0.198542</td>\n",
       "      <td>1.820618</td>\n",
       "      <td>-0.074209</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>198363</th>\n",
       "      <td>-1.135983</td>\n",
       "      <td>-2.150924</td>\n",
       "      <td>-1.078870</td>\n",
       "      <td>-0.922886</td>\n",
       "      <td>-1.936992</td>\n",
       "      <td>0.819695</td>\n",
       "      <td>3.101553</td>\n",
       "      <td>-0.311864</td>\n",
       "      <td>-1.447371</td>\n",
       "      <td>-0.418445</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>198364</th>\n",
       "      <td>-0.664842</td>\n",
       "      <td>0.635278</td>\n",
       "      <td>1.398580</td>\n",
       "      <td>-0.412006</td>\n",
       "      <td>0.357441</td>\n",
       "      <td>-0.090934</td>\n",
       "      <td>0.804617</td>\n",
       "      <td>0.076064</td>\n",
       "      <td>-0.906568</td>\n",
       "      <td>-0.463076</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>198365 rows × 54 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              V1        V2        V3        V4        V5        V6        V7  \\\n",
       "0      -1.359807 -0.072781  2.536347  1.378155 -0.338321  0.462388  0.239599   \n",
       "1       1.191857  0.266151  0.166480  0.448154  0.060018 -0.082361 -0.078803   \n",
       "2      -1.358354 -1.340163  1.773209  0.379780 -0.503198  1.800499  0.791461   \n",
       "3      -0.966272 -0.185226  1.792993 -0.863291 -0.010309  1.247203  0.237609   \n",
       "4      -1.158233  0.877737  1.548718  0.403034 -0.407193  0.095921  0.592941   \n",
       "...          ...       ...       ...       ...       ...       ...       ...   \n",
       "198360 -0.474163  1.438815 -0.816746  0.786791  1.067826 -0.207767  1.251817   \n",
       "198361  1.501100 -2.329402 -0.028328 -0.579340 -2.269449 -0.208058 -1.307538   \n",
       "198362  1.957960 -0.677669 -0.395253  0.762448 -0.685747  0.201289 -0.854448   \n",
       "198363 -1.135983 -2.150924 -1.078870 -0.922886 -1.936992  0.819695  3.101553   \n",
       "198364 -0.664842  0.635278  1.398580 -0.412006  0.357441 -0.090934  0.804617   \n",
       "\n",
       "              V8        V9       V10  ...  14  15  16  17  18  19  20  21  22  \\\n",
       "0       0.098698  0.363787  0.090794  ...   0   0   0   0   0   0   0   0   0   \n",
       "1       0.085102 -0.255425 -0.166974  ...   0   0   0   0   0   0   0   0   0   \n",
       "2       0.247676 -1.514654  0.207643  ...   0   0   0   0   0   0   0   0   0   \n",
       "3       0.377436 -1.387024 -0.054952  ...   0   0   0   0   0   0   0   0   0   \n",
       "4      -0.270533  0.817739  0.753074  ...   0   0   0   0   0   0   0   0   0   \n",
       "...          ...       ...       ...  ...  ..  ..  ..  ..  ..  ..  ..  ..  ..   \n",
       "198360  0.140304 -0.903642 -0.286039  ...   0   0   0   0   0   0   0   0   0   \n",
       "198361  0.083778  0.774633  0.584604  ...   0   0   0   0   0   0   0   0   0   \n",
       "198362  0.198542  1.820618 -0.074209  ...   0   0   0   0   0   0   0   0   0   \n",
       "198363 -0.311864 -1.447371 -0.418445  ...   0   0   0   0   0   0   0   0   0   \n",
       "198364  0.076064 -0.906568 -0.463076  ...   0   0   0   0   0   0   0   0   0   \n",
       "\n",
       "        23  \n",
       "0        0  \n",
       "1        0  \n",
       "2        0  \n",
       "3        0  \n",
       "4        0  \n",
       "...     ..  \n",
       "198360   0  \n",
       "198361   0  \n",
       "198362   0  \n",
       "198363   0  \n",
       "198364   0  \n",
       "\n",
       "[198365 rows x 54 columns]"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#lets get dummies\n",
    "T_dummies = pd.get_dummies(data['HourBank'])\n",
    "\n",
    "#lets merge it\n",
    "data = pd.concat([data, T_dummies], axis = 1)\n",
    "\n",
    "# then drop the redundant column\n",
    "data = data.drop(['HourBank','ID','Time'], axis = 1)\n",
    "\n",
    "\n",
    "##### Repeat the same for test data\n",
    "\n",
    "Tt_dummies = pd.get_dummies(test['HourBank'])\n",
    "a = list(Tt_dummies.columns.values)\n",
    "b = [0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22, 23]\n",
    "c = [x for x in b if x not in a]\n",
    "Tt_dummies = pd.concat([Tt_dummies, pd.DataFrame(columns = c)]).fillna(0)  \n",
    "#lets merge it\n",
    "test = pd.concat([test, Tt_dummies], axis = 1)\n",
    "# then drop the redundant column\n",
    "test = test.drop(['HourBank','ID','Time'], axis = 1)\n",
    "\n",
    "\n",
    "data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "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>V1</th>\n",
       "      <th>V2</th>\n",
       "      <th>V3</th>\n",
       "      <th>V4</th>\n",
       "      <th>V5</th>\n",
       "      <th>V6</th>\n",
       "      <th>V7</th>\n",
       "      <th>V8</th>\n",
       "      <th>V9</th>\n",
       "      <th>V10</th>\n",
       "      <th>...</th>\n",
       "      <th>14</th>\n",
       "      <th>15</th>\n",
       "      <th>16</th>\n",
       "      <th>17</th>\n",
       "      <th>18</th>\n",
       "      <th>19</th>\n",
       "      <th>20</th>\n",
       "      <th>21</th>\n",
       "      <th>22</th>\n",
       "      <th>23</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2.079192</td>\n",
       "      <td>0.052662</td>\n",
       "      <td>-1.755167</td>\n",
       "      <td>0.418802</td>\n",
       "      <td>0.310472</td>\n",
       "      <td>-0.886053</td>\n",
       "      <td>0.083326</td>\n",
       "      <td>-0.149153</td>\n",
       "      <td>0.721121</td>\n",
       "      <td>-0.334404</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.596563</td>\n",
       "      <td>-1.588177</td>\n",
       "      <td>-1.186718</td>\n",
       "      <td>-1.710578</td>\n",
       "      <td>-0.378661</td>\n",
       "      <td>0.906028</td>\n",
       "      <td>-0.796517</td>\n",
       "      <td>0.443712</td>\n",
       "      <td>2.149723</td>\n",
       "      <td>-0.903409</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2.273454</td>\n",
       "      <td>-1.447207</td>\n",
       "      <td>-0.946779</td>\n",
       "      <td>-1.738063</td>\n",
       "      <td>-1.074172</td>\n",
       "      <td>-0.139592</td>\n",
       "      <td>-1.266807</td>\n",
       "      <td>-0.047025</td>\n",
       "      <td>-1.345062</td>\n",
       "      <td>1.713041</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.987406</td>\n",
       "      <td>0.140840</td>\n",
       "      <td>-2.001286</td>\n",
       "      <td>1.071763</td>\n",
       "      <td>0.717042</td>\n",
       "      <td>-0.880590</td>\n",
       "      <td>0.670095</td>\n",
       "      <td>-0.288313</td>\n",
       "      <td>-0.229202</td>\n",
       "      <td>0.553870</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.398751</td>\n",
       "      <td>1.370313</td>\n",
       "      <td>-0.715049</td>\n",
       "      <td>-0.867699</td>\n",
       "      <td>1.157188</td>\n",
       "      <td>-0.364107</td>\n",
       "      <td>1.151124</td>\n",
       "      <td>-0.316554</td>\n",
       "      <td>0.756304</td>\n",
       "      <td>0.843054</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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",
       "      <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>86437</th>\n",
       "      <td>-11.881118</td>\n",
       "      <td>10.071785</td>\n",
       "      <td>-9.834783</td>\n",
       "      <td>-2.066656</td>\n",
       "      <td>-5.364473</td>\n",
       "      <td>-2.606837</td>\n",
       "      <td>-4.918215</td>\n",
       "      <td>7.305334</td>\n",
       "      <td>1.914428</td>\n",
       "      <td>4.356170</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>86438</th>\n",
       "      <td>-0.732789</td>\n",
       "      <td>-0.055080</td>\n",
       "      <td>2.035030</td>\n",
       "      <td>-0.738589</td>\n",
       "      <td>0.868229</td>\n",
       "      <td>1.058415</td>\n",
       "      <td>0.024330</td>\n",
       "      <td>0.294869</td>\n",
       "      <td>0.584800</td>\n",
       "      <td>-0.975926</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>86439</th>\n",
       "      <td>1.919565</td>\n",
       "      <td>-0.301254</td>\n",
       "      <td>-3.249640</td>\n",
       "      <td>-0.557828</td>\n",
       "      <td>2.630515</td>\n",
       "      <td>3.031260</td>\n",
       "      <td>-0.296827</td>\n",
       "      <td>0.708417</td>\n",
       "      <td>0.432454</td>\n",
       "      <td>-0.484782</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>86440</th>\n",
       "      <td>-0.240440</td>\n",
       "      <td>0.530483</td>\n",
       "      <td>0.702510</td>\n",
       "      <td>0.689799</td>\n",
       "      <td>-0.377961</td>\n",
       "      <td>0.623708</td>\n",
       "      <td>-0.686180</td>\n",
       "      <td>0.679145</td>\n",
       "      <td>0.392087</td>\n",
       "      <td>-0.399126</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>86441</th>\n",
       "      <td>-0.533413</td>\n",
       "      <td>-0.189733</td>\n",
       "      <td>0.703337</td>\n",
       "      <td>-0.506271</td>\n",
       "      <td>-0.012546</td>\n",
       "      <td>-0.649617</td>\n",
       "      <td>1.577006</td>\n",
       "      <td>-0.414650</td>\n",
       "      <td>0.486180</td>\n",
       "      <td>-0.915427</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>86442 rows × 53 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              V1         V2        V3        V4        V5        V6        V7  \\\n",
       "0       2.079192   0.052662 -1.755167  0.418802  0.310472 -0.886053  0.083326   \n",
       "1       1.596563  -1.588177 -1.186718 -1.710578 -0.378661  0.906028 -0.796517   \n",
       "2       2.273454  -1.447207 -0.946779 -1.738063 -1.074172 -0.139592 -1.266807   \n",
       "3       1.987406   0.140840 -2.001286  1.071763  0.717042 -0.880590  0.670095   \n",
       "4      -0.398751   1.370313 -0.715049 -0.867699  1.157188 -0.364107  1.151124   \n",
       "...          ...        ...       ...       ...       ...       ...       ...   \n",
       "86437 -11.881118  10.071785 -9.834783 -2.066656 -5.364473 -2.606837 -4.918215   \n",
       "86438  -0.732789  -0.055080  2.035030 -0.738589  0.868229  1.058415  0.024330   \n",
       "86439   1.919565  -0.301254 -3.249640 -0.557828  2.630515  3.031260 -0.296827   \n",
       "86440  -0.240440   0.530483  0.702510  0.689799 -0.377961  0.623708 -0.686180   \n",
       "86441  -0.533413  -0.189733  0.703337 -0.506271 -0.012546 -0.649617  1.577006   \n",
       "\n",
       "             V8        V9       V10  ...   14   15   16   17   18   19   20  \\\n",
       "0     -0.149153  0.721121 -0.334404  ...  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "1      0.443712  2.149723 -0.903409  ...  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "2     -0.047025 -1.345062  1.713041  ...  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "3     -0.288313 -0.229202  0.553870  ...  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "4     -0.316554  0.756304  0.843054  ...  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "...         ...       ...       ...  ...  ...  ...  ...  ...  ...  ...  ...   \n",
       "86437  7.305334  1.914428  4.356170  ...  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "86438  0.294869  0.584800 -0.975926  ...  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "86439  0.708417  0.432454 -0.484782  ...  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "86440  0.679145  0.392087 -0.399126  ...  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "86441 -0.414650  0.486180 -0.915427  ...  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "\n",
       "        21   22   23  \n",
       "0      0.0  0.0  0.0  \n",
       "1      0.0  0.0  0.0  \n",
       "2      0.0  0.0  0.0  \n",
       "3      0.0  0.0  0.0  \n",
       "4      0.0  0.0  0.0  \n",
       "...    ...  ...  ...  \n",
       "86437  0.0  0.0  1.0  \n",
       "86438  0.0  0.0  1.0  \n",
       "86439  0.0  0.0  1.0  \n",
       "86440  0.0  0.0  1.0  \n",
       "86441  0.0  0.0  1.0  \n",
       "\n",
       "[86442 rows x 53 columns]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Shuffle Dataset "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "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>V1</th>\n",
       "      <th>V2</th>\n",
       "      <th>V3</th>\n",
       "      <th>V4</th>\n",
       "      <th>V5</th>\n",
       "      <th>V6</th>\n",
       "      <th>V7</th>\n",
       "      <th>V8</th>\n",
       "      <th>V9</th>\n",
       "      <th>V10</th>\n",
       "      <th>...</th>\n",
       "      <th>14</th>\n",
       "      <th>15</th>\n",
       "      <th>16</th>\n",
       "      <th>17</th>\n",
       "      <th>18</th>\n",
       "      <th>19</th>\n",
       "      <th>20</th>\n",
       "      <th>21</th>\n",
       "      <th>22</th>\n",
       "      <th>23</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.158608</td>\n",
       "      <td>-0.682005</td>\n",
       "      <td>0.834863</td>\n",
       "      <td>-0.860150</td>\n",
       "      <td>-1.252565</td>\n",
       "      <td>-0.105402</td>\n",
       "      <td>-0.972816</td>\n",
       "      <td>0.436383</td>\n",
       "      <td>2.047848</td>\n",
       "      <td>-0.860108</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.056601</td>\n",
       "      <td>0.125731</td>\n",
       "      <td>1.338773</td>\n",
       "      <td>2.673969</td>\n",
       "      <td>-0.805437</td>\n",
       "      <td>0.219487</td>\n",
       "      <td>-0.557807</td>\n",
       "      <td>0.287495</td>\n",
       "      <td>-0.102484</td>\n",
       "      <td>0.719404</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.190746</td>\n",
       "      <td>1.265327</td>\n",
       "      <td>-1.477330</td>\n",
       "      <td>1.578233</td>\n",
       "      <td>0.805379</td>\n",
       "      <td>-1.620875</td>\n",
       "      <td>0.685694</td>\n",
       "      <td>-0.259457</td>\n",
       "      <td>-0.738327</td>\n",
       "      <td>-1.385118</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-4.312479</td>\n",
       "      <td>1.886476</td>\n",
       "      <td>-2.338634</td>\n",
       "      <td>-0.475243</td>\n",
       "      <td>-1.185444</td>\n",
       "      <td>-2.112079</td>\n",
       "      <td>-2.122793</td>\n",
       "      <td>0.272565</td>\n",
       "      <td>0.290273</td>\n",
       "      <td>-3.833741</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.887287</td>\n",
       "      <td>1.390002</td>\n",
       "      <td>1.219686</td>\n",
       "      <td>1.661425</td>\n",
       "      <td>1.009228</td>\n",
       "      <td>-0.733908</td>\n",
       "      <td>0.855829</td>\n",
       "      <td>0.000077</td>\n",
       "      <td>-1.275631</td>\n",
       "      <td>-0.433394</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</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",
       "      <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>761</th>\n",
       "      <td>-5.791495</td>\n",
       "      <td>-5.243817</td>\n",
       "      <td>1.800729</td>\n",
       "      <td>2.720080</td>\n",
       "      <td>2.840993</td>\n",
       "      <td>-1.873182</td>\n",
       "      <td>-1.987863</td>\n",
       "      <td>0.824355</td>\n",
       "      <td>-0.081015</td>\n",
       "      <td>-0.036398</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>762</th>\n",
       "      <td>1.113731</td>\n",
       "      <td>-0.038003</td>\n",
       "      <td>1.504975</td>\n",
       "      <td>1.473218</td>\n",
       "      <td>-1.270853</td>\n",
       "      <td>-0.519113</td>\n",
       "      <td>-0.565426</td>\n",
       "      <td>0.070582</td>\n",
       "      <td>0.896623</td>\n",
       "      <td>-0.194678</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>763</th>\n",
       "      <td>-2.401980</td>\n",
       "      <td>-1.684031</td>\n",
       "      <td>2.238891</td>\n",
       "      <td>0.577867</td>\n",
       "      <td>-1.270153</td>\n",
       "      <td>0.867902</td>\n",
       "      <td>1.337289</td>\n",
       "      <td>0.380959</td>\n",
       "      <td>-0.400667</td>\n",
       "      <td>-1.088591</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>764</th>\n",
       "      <td>1.799034</td>\n",
       "      <td>0.440732</td>\n",
       "      <td>-1.200828</td>\n",
       "      <td>3.405403</td>\n",
       "      <td>1.675520</td>\n",
       "      <td>2.111250</td>\n",
       "      <td>-0.164396</td>\n",
       "      <td>0.412595</td>\n",
       "      <td>-0.014265</td>\n",
       "      <td>1.210987</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>765</th>\n",
       "      <td>1.166597</td>\n",
       "      <td>-0.155948</td>\n",
       "      <td>0.434363</td>\n",
       "      <td>0.384343</td>\n",
       "      <td>-0.427271</td>\n",
       "      <td>-0.132608</td>\n",
       "      <td>-0.168273</td>\n",
       "      <td>-0.006811</td>\n",
       "      <td>0.341815</td>\n",
       "      <td>-0.172891</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>766 rows × 54 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           V1        V2        V3        V4        V5        V6        V7  \\\n",
       "0    1.158608 -0.682005  0.834863 -0.860150 -1.252565 -0.105402 -0.972816   \n",
       "1    1.056601  0.125731  1.338773  2.673969 -0.805437  0.219487 -0.557807   \n",
       "2    1.190746  1.265327 -1.477330  1.578233  0.805379 -1.620875  0.685694   \n",
       "3   -4.312479  1.886476 -2.338634 -0.475243 -1.185444 -2.112079 -2.122793   \n",
       "4   -0.887287  1.390002  1.219686  1.661425  1.009228 -0.733908  0.855829   \n",
       "..        ...       ...       ...       ...       ...       ...       ...   \n",
       "761 -5.791495 -5.243817  1.800729  2.720080  2.840993 -1.873182 -1.987863   \n",
       "762  1.113731 -0.038003  1.504975  1.473218 -1.270853 -0.519113 -0.565426   \n",
       "763 -2.401980 -1.684031  2.238891  0.577867 -1.270153  0.867902  1.337289   \n",
       "764  1.799034  0.440732 -1.200828  3.405403  1.675520  2.111250 -0.164396   \n",
       "765  1.166597 -0.155948  0.434363  0.384343 -0.427271 -0.132608 -0.168273   \n",
       "\n",
       "           V8        V9       V10  ...  14  15  16  17  18  19  20  21  22  23  \n",
       "0    0.436383  2.047848 -0.860108  ...   0   0   0   0   0   0   0   1   0   0  \n",
       "1    0.287495 -0.102484  0.719404  ...   0   0   1   0   0   0   0   0   0   0  \n",
       "2   -0.259457 -0.738327 -1.385118  ...   0   0   0   0   0   0   0   1   0   0  \n",
       "3    0.272565  0.290273 -3.833741  ...   0   0   0   0   0   0   0   1   0   0  \n",
       "4    0.000077 -1.275631 -0.433394  ...   0   0   0   0   0   0   0   0   0   0  \n",
       "..        ...       ...       ...  ...  ..  ..  ..  ..  ..  ..  ..  ..  ..  ..  \n",
       "761  0.824355 -0.081015 -0.036398  ...   0   0   0   0   0   0   0   0   0   0  \n",
       "762  0.070582  0.896623 -0.194678  ...   0   0   0   0   0   0   0   1   0   0  \n",
       "763  0.380959 -0.400667 -1.088591  ...   0   0   1   0   0   0   0   0   0   0  \n",
       "764  0.412595 -0.014265  1.210987  ...   0   0   0   0   0   0   0   0   0   0  \n",
       "765 -0.006811  0.341815 -0.172891  ...   0   0   0   0   0   0   0   0   0   0  \n",
       "\n",
       "[766 rows x 54 columns]"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Balance the data based on column class\n",
    "g = data.groupby('Class')\n",
    "g = pd.DataFrame(g.apply(lambda x: x.sample(g.size().min()).reset_index(drop=True)))\n",
    "g = g.reset_index(drop=True)\n",
    "\n",
    "###shuffle dataset\n",
    "g = g.sample(frac=1).reset_index(drop=True)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Balance and Standardize the inputs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {},
   "outputs": [],
   "source": [
    "# The inputs are all columns in the csv, except for the first one [:,0]\n",
    "# (which is just the arbitrary customer IDs that bear no useful information),\n",
    "# and the last one [:,-1] (which is our targets)\n",
    "balanced_inputs = g.drop(['Class'],axis=1)\n",
    "# The targets are in the last column. That's how datasets are conventionally organized.\n",
    "balanced_targets = g['Class'].astype(np.int)\n",
    "unbalanced_inputs = data.drop(['Class'],axis=1)\n",
    "# The targets are in the last column. That's how datasets are conventionally organized.\n",
    "unbalanced_targets = data['Class'].astype(np.int)\n",
    "\n",
    "balanced_inputs = preprocessing.scale(balanced_inputs)\n",
    "scaled_unbalanced_inputs = preprocessing.scale(unbalanced_inputs)\n",
    "\n",
    "test_inputsx = preprocessing.scale(test)\n",
    "\n",
    "shuffled_inputs = balanced_inputs\n",
    "shuffled_targets = balanced_targets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "309 612 0.5049019607843137\n",
      "37 76 0.4868421052631579\n",
      "37 78 0.47435897435897434\n"
     ]
    }
   ],
   "source": [
    "# Count the total number of samples\n",
    "samples_count = shuffled_inputs.shape[0]\n",
    "\n",
    "# Count the samples in each subset, assuming we want 80-10-10 distribution of training, validation, and test.\n",
    "# Naturally, the numbers are integers.\n",
    "train_samples_count = int(0.8 * samples_count)\n",
    "validation_samples_count = int(0.1 * samples_count)\n",
    "\n",
    "# The 'test' dataset contains all remaining data.\n",
    "test_samples_count = samples_count - train_samples_count - validation_samples_count\n",
    "\n",
    "# Create variables that record the inputs and targets for training\n",
    "# In our shuffled dataset, they are the first \"train_samples_count\" observations\n",
    "train_inputs = shuffled_inputs[:train_samples_count]\n",
    "train_targets = shuffled_targets[:train_samples_count]\n",
    "\n",
    "# Create variables that record the inputs and targets for validation.\n",
    "# They are the next \"validation_samples_count\" observations, folllowing the \"train_samples_count\" we already assigned\n",
    "validation_inputs = shuffled_inputs[train_samples_count:train_samples_count+validation_samples_count]\n",
    "validation_targets = shuffled_targets[train_samples_count:train_samples_count+validation_samples_count]\n",
    "\n",
    "# Create variables that record the inputs and targets for test.\n",
    "# They are everything that is remaining.\n",
    "test_inputs = shuffled_inputs[train_samples_count+validation_samples_count:]\n",
    "test_targets = shuffled_targets[train_samples_count+validation_samples_count:]\n",
    "\n",
    "# We balanced our dataset to be 50-50 (for targets 0 and 1), but the training, validation, and test were \n",
    "# taken from a shuffled dataset. Check if they are balanced, too. Note that each time you rerun this code, \n",
    "# you will get different values, as each time they are shuffled randomly.\n",
    "# Normally you preprocess ONCE, so you need not rerun this code once it is done.\n",
    "# If you rerun this whole sheet, the npzs will be overwritten with your newly preprocessed data.\n",
    "\n",
    "# Print the number of targets that are 1s, the total number of samples, and the proportion for training, validation, and test.\n",
    "print(np.sum(train_targets), train_samples_count, np.sum(train_targets) / train_samples_count)\n",
    "print(np.sum(validation_targets), validation_samples_count, np.sum(validation_targets) / validation_samples_count)\n",
    "print(np.sum(test_targets), test_samples_count, np.sum(test_targets) / test_samples_count)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Split the unbalanced dataset into train,  and test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Unbalanced datset\n",
    "train_test_split(scaled_unbalanced_inputs, unbalanced_targets)\n",
    "# declare 4 variables for the split\n",
    "x_train1, x_test1, y_train1, y_test1 = train_test_split(scaled_unbalanced_inputs, unbalanced_targets, #train_size = 0.75, \n",
    "                                                                            test_size = 0.25, random_state = 20)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Quick modelling using Default parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Now, we will create an array of Classifiers and append different classification models to our array\n",
    "classifiers = [] \n",
    "\n",
    "mod1 = xgboost.XGBClassifier()\n",
    "classifiers.append(mod1)\n",
    "mod2 = svm.SVC()\n",
    "classifiers.append(mod2)\n",
    "mod3 = RandomForestClassifier()\n",
    "classifiers.append(mod3)\n",
    "mod4 = LogisticRegression()\n",
    "classifiers.append(mod4)\n",
    "mod5 = KNeighborsClassifier(3)\n",
    "classifiers.append(mod5)\n",
    "mod6 = AdaBoostClassifier()\n",
    "classifiers.append(mod6)\n",
    "mod7= GaussianNB()\n",
    "classifiers.append(mod7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "XGBClassifier(base_score=0.5, booster='gbtree', colsample_bylevel=1,\n",
      "              colsample_bynode=1, colsample_bytree=1, gamma=0, gpu_id=-1,\n",
      "              importance_type='gain', interaction_constraints='',\n",
      "              learning_rate=0.300000012, max_delta_step=0, max_depth=6,\n",
      "              min_child_weight=1, missing=nan, monotone_constraints='()',\n",
      "              n_estimators=100, n_jobs=0, num_parallel_tree=1,\n",
      "              objective='binary:logistic', random_state=0, reg_alpha=0,\n",
      "              reg_lambda=1, scale_pos_weight=1, subsample=1,\n",
      "              tree_method='exact', validate_parameters=1, verbosity=None)\n",
      "Accuracy of trainset XGBClassifier is 1.0\n",
      "Accuracy of testset XGBClassifier is 0.8974358974358975\n",
      "Confusion Matrix of testset XGBClassifier is [[37  4]\n",
      " [ 4 33]]\n",
      "SVC(C=1.0, break_ties=False, cache_size=200, class_weight=None, coef0=0.0,\n",
      "    decision_function_shape='ovr', degree=3, gamma='scale', kernel='rbf',\n",
      "    max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
      "    tol=0.001, verbose=False)\n",
      "Accuracy of trainset SVC is 0.9428104575163399\n",
      "Accuracy of testset SVC is 0.9230769230769231\n",
      "Confusion Matrix of testset SVC is [[40  1]\n",
      " [ 5 32]]\n",
      "RandomForestClassifier(bootstrap=True, ccp_alpha=0.0, class_weight=None,\n",
      "                       criterion='gini', max_depth=None, max_features='auto',\n",
      "                       max_leaf_nodes=None, max_samples=None,\n",
      "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
      "                       min_samples_leaf=1, min_samples_split=2,\n",
      "                       min_weight_fraction_leaf=0.0, n_estimators=100,\n",
      "                       n_jobs=None, oob_score=False, random_state=None,\n",
      "                       verbose=0, warm_start=False)\n",
      "Accuracy of trainset RandomForestClassifier is 1.0\n",
      "Accuracy of testset RandomForestClassifier is 0.9487179487179487\n",
      "Confusion Matrix of testset RandomForestClassifier is [[41  0]\n",
      " [ 4 33]]\n",
      "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
      "                   intercept_scaling=1, l1_ratio=None, max_iter=100,\n",
      "                   multi_class='auto', n_jobs=None, penalty='l2',\n",
      "                   random_state=None, solver='lbfgs', tol=0.0001, verbose=0,\n",
      "                   warm_start=False)\n",
      "Accuracy of trainset LogisticRegression is 0.9673202614379085\n",
      "Accuracy of testset LogisticRegression is 0.9230769230769231\n",
      "Confusion Matrix of testset LogisticRegression is [[39  2]\n",
      " [ 4 33]]\n",
      "KNeighborsClassifier(algorithm='auto', leaf_size=30, metric='minkowski',\n",
      "                     metric_params=None, n_jobs=None, n_neighbors=3, p=2,\n",
      "                     weights='uniform')\n",
      "Accuracy of trainset KNeighborsClassifier is 0.9248366013071896\n",
      "Accuracy of testset KNeighborsClassifier is 0.8717948717948718\n",
      "Confusion Matrix of testset KNeighborsClassifier is [[41  0]\n",
      " [10 27]]\n",
      "AdaBoostClassifier(algorithm='SAMME.R', base_estimator=None, learning_rate=1.0,\n",
      "                   n_estimators=50, random_state=None)\n",
      "Accuracy of trainset AdaBoostClassifier is 0.9967320261437909\n",
      "Accuracy of testset AdaBoostClassifier is 0.9358974358974359\n",
      "Confusion Matrix of testset AdaBoostClassifier is [[39  2]\n",
      " [ 3 34]]\n",
      "GaussianNB(priors=None, var_smoothing=1e-09)\n",
      "Accuracy of trainset GaussianNB is 0.9117647058823529\n",
      "Accuracy of testset GaussianNB is 0.8846153846153846\n",
      "Confusion Matrix of testset GaussianNB is [[38  3]\n",
      " [ 6 31]]\n"
     ]
    }
   ],
   "source": [
    "#Lets fit the models into anarray\n",
    "\n",
    "for clf in classifiers:\n",
    "    clf.fit(train_inputs,train_targets)\n",
    "    y_pred= clf.predict(test_inputs)\n",
    "    y_tr = clf.predict(train_inputs)\n",
    "    acc_tr = accuracy_score(train_targets, y_tr)\n",
    "    acc = accuracy_score(test_targets, y_pred)\n",
    "    mn = type(clf).__name__\n",
    "    \n",
    "    print(clf)\n",
    "    print(\"Accuracy of trainset %s is %s\"%(mn, acc_tr))\n",
    "    print(\"Accuracy of testset %s is %s\"%(mn, acc))\n",
    "    cm = confusion_matrix(test_targets, y_pred)\n",
    "    print(\"Confusion Matrix of testset %s is %s\"%(mn, cm))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Deep Learning Modelling\n",
    "Outline, optimizers, loss, early stopping and training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "# convert all value into array\n",
    "validation_inputs = np.array(validation_inputs)\n",
    "validation_targets = np.array(validation_targets)\n",
    "train_targets = np.array(train_targets)\n",
    "train_inputs = np.array(train_inputs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 612 samples, validate on 76 samples\n",
      "Epoch 1/100\n",
      "612/612 - 1s - loss: 0.6909 - accuracy: 0.5196 - val_loss: 0.6889 - val_accuracy: 0.6316\n",
      "Epoch 2/100\n",
      "612/612 - 0s - loss: 0.6881 - accuracy: 0.6618 - val_loss: 0.6871 - val_accuracy: 0.6053\n",
      "Epoch 3/100\n",
      "612/612 - 0s - loss: 0.6849 - accuracy: 0.6536 - val_loss: 0.6840 - val_accuracy: 0.6842\n",
      "Epoch 4/100\n",
      "612/612 - 0s - loss: 0.6810 - accuracy: 0.6732 - val_loss: 0.6806 - val_accuracy: 0.6842\n",
      "Epoch 5/100\n",
      "612/612 - 0s - loss: 0.6758 - accuracy: 0.6977 - val_loss: 0.6768 - val_accuracy: 0.6711\n",
      "Epoch 6/100\n",
      "612/612 - 0s - loss: 0.6689 - accuracy: 0.7206 - val_loss: 0.6710 - val_accuracy: 0.6842\n",
      "Epoch 7/100\n",
      "612/612 - 0s - loss: 0.6590 - accuracy: 0.7386 - val_loss: 0.6639 - val_accuracy: 0.7368\n",
      "Epoch 8/100\n",
      "612/612 - 0s - loss: 0.6462 - accuracy: 0.7533 - val_loss: 0.6541 - val_accuracy: 0.7500\n",
      "Epoch 9/100\n",
      "612/612 - 0s - loss: 0.6308 - accuracy: 0.7696 - val_loss: 0.6416 - val_accuracy: 0.7500\n",
      "Epoch 10/100\n",
      "612/612 - 0s - loss: 0.6132 - accuracy: 0.7958 - val_loss: 0.6266 - val_accuracy: 0.7763\n",
      "Epoch 11/100\n",
      "612/612 - 0s - loss: 0.5917 - accuracy: 0.8121 - val_loss: 0.6109 - val_accuracy: 0.8026\n",
      "Epoch 12/100\n",
      "612/612 - 0s - loss: 0.5693 - accuracy: 0.8186 - val_loss: 0.5940 - val_accuracy: 0.7763\n",
      "Epoch 13/100\n",
      "612/612 - 0s - loss: 0.5442 - accuracy: 0.8203 - val_loss: 0.5760 - val_accuracy: 0.7895\n",
      "Epoch 14/100\n",
      "612/612 - 0s - loss: 0.5188 - accuracy: 0.8366 - val_loss: 0.5585 - val_accuracy: 0.8026\n",
      "Epoch 15/100\n",
      "612/612 - 0s - loss: 0.4928 - accuracy: 0.8415 - val_loss: 0.5422 - val_accuracy: 0.8026\n",
      "Epoch 16/100\n",
      "612/612 - 0s - loss: 0.4682 - accuracy: 0.8513 - val_loss: 0.5259 - val_accuracy: 0.8158\n",
      "Epoch 17/100\n",
      "612/612 - 0s - loss: 0.4440 - accuracy: 0.8660 - val_loss: 0.5095 - val_accuracy: 0.8158\n",
      "Epoch 18/100\n",
      "612/612 - 0s - loss: 0.4204 - accuracy: 0.8742 - val_loss: 0.4937 - val_accuracy: 0.8289\n",
      "Epoch 19/100\n",
      "612/612 - 0s - loss: 0.3988 - accuracy: 0.8758 - val_loss: 0.4796 - val_accuracy: 0.8289\n",
      "Epoch 20/100\n",
      "612/612 - 0s - loss: 0.3778 - accuracy: 0.8889 - val_loss: 0.4656 - val_accuracy: 0.8289\n",
      "Epoch 21/100\n",
      "612/612 - 0s - loss: 0.3596 - accuracy: 0.8954 - val_loss: 0.4538 - val_accuracy: 0.8289\n",
      "Epoch 22/100\n",
      "612/612 - 0s - loss: 0.3426 - accuracy: 0.8987 - val_loss: 0.4412 - val_accuracy: 0.8289\n",
      "Epoch 23/100\n",
      "612/612 - 0s - loss: 0.3273 - accuracy: 0.8987 - val_loss: 0.4280 - val_accuracy: 0.8289\n",
      "Epoch 24/100\n",
      "612/612 - 0s - loss: 0.3135 - accuracy: 0.9036 - val_loss: 0.4173 - val_accuracy: 0.8421\n",
      "Epoch 25/100\n",
      "612/612 - 0s - loss: 0.3008 - accuracy: 0.9085 - val_loss: 0.4072 - val_accuracy: 0.8421\n",
      "Epoch 26/100\n",
      "612/612 - 0s - loss: 0.2888 - accuracy: 0.9085 - val_loss: 0.3991 - val_accuracy: 0.8553\n",
      "Epoch 27/100\n",
      "612/612 - 0s - loss: 0.2779 - accuracy: 0.9118 - val_loss: 0.3935 - val_accuracy: 0.8553\n",
      "Epoch 28/100\n",
      "612/612 - 0s - loss: 0.2679 - accuracy: 0.9134 - val_loss: 0.3852 - val_accuracy: 0.8553\n",
      "Epoch 29/100\n",
      "612/612 - 0s - loss: 0.2565 - accuracy: 0.9167 - val_loss: 0.3719 - val_accuracy: 0.8816\n",
      "Epoch 30/100\n",
      "612/612 - 0s - loss: 0.2457 - accuracy: 0.9216 - val_loss: 0.3580 - val_accuracy: 0.8816\n",
      "Epoch 31/100\n",
      "612/612 - 0s - loss: 0.2363 - accuracy: 0.9216 - val_loss: 0.3424 - val_accuracy: 0.8816\n",
      "Epoch 32/100\n",
      "612/612 - 0s - loss: 0.2273 - accuracy: 0.9265 - val_loss: 0.3328 - val_accuracy: 0.8947\n",
      "Epoch 33/100\n",
      "612/612 - 0s - loss: 0.2198 - accuracy: 0.9297 - val_loss: 0.3234 - val_accuracy: 0.8947\n",
      "Epoch 34/100\n",
      "612/612 - 0s - loss: 0.2127 - accuracy: 0.9330 - val_loss: 0.3172 - val_accuracy: 0.8947\n",
      "Epoch 35/100\n",
      "612/612 - 0s - loss: 0.2064 - accuracy: 0.9281 - val_loss: 0.3162 - val_accuracy: 0.8947\n",
      "Epoch 36/100\n",
      "612/612 - 0s - loss: 0.1991 - accuracy: 0.9297 - val_loss: 0.3045 - val_accuracy: 0.8947\n",
      "Epoch 37/100\n",
      "612/612 - 0s - loss: 0.1936 - accuracy: 0.9314 - val_loss: 0.2979 - val_accuracy: 0.8947\n",
      "Epoch 38/100\n",
      "612/612 - 0s - loss: 0.1881 - accuracy: 0.9363 - val_loss: 0.2943 - val_accuracy: 0.8947\n",
      "Epoch 39/100\n",
      "612/612 - 0s - loss: 0.1827 - accuracy: 0.9379 - val_loss: 0.2885 - val_accuracy: 0.8816\n",
      "Epoch 40/100\n",
      "612/612 - 0s - loss: 0.1770 - accuracy: 0.9428 - val_loss: 0.2885 - val_accuracy: 0.8816\n",
      "Epoch 41/100\n",
      "612/612 - 0s - loss: 0.1710 - accuracy: 0.9444 - val_loss: 0.2832 - val_accuracy: 0.8947\n",
      "Epoch 42/100\n",
      "612/612 - 0s - loss: 0.1656 - accuracy: 0.9444 - val_loss: 0.2841 - val_accuracy: 0.8947\n",
      "Epoch 43/100\n",
      "612/612 - 0s - loss: 0.1595 - accuracy: 0.9461 - val_loss: 0.2879 - val_accuracy: 0.9079\n"
     ]
    }
   ],
   "source": [
    "# Set the input and output sizes\n",
    "input_size = 53\n",
    "output_size = 2\n",
    "# Use same hidden layer size for both hidden layers. Not a necessity.\n",
    "hidden_layer_size = 6\n",
    "    \n",
    "# define how the model will look like\n",
    "model = tf.keras.Sequential([\n",
    "    # tf.keras.layers.Dense is basically implementing: output = activation(dot(input, weight) + bias)\n",
    "    # it takes several arguments, but the most important ones for us are the hidden_layer_size and the activation function\n",
    "    tf.keras.layers.Dense(hidden_layer_size, activation='relu'), # 1st hidden layer\n",
    "    tf.keras.layers.Dense(hidden_layer_size, activation='relu'), # 2nd hidden layer\n",
    "    tf.keras.layers.Dense(hidden_layer_size, activation='relu'), # 2nd hidden layer\n",
    "    tf.keras.layers.Dense(hidden_layer_size, activation='relu'), # 2nd hidden layer\n",
    "    tf.keras.layers.Dense(hidden_layer_size, activation='relu'), # 2nd hidden layer\n",
    "    # the final layer is no different, we just make sure to activate it with softmax\n",
    "    tf.keras.layers.Dense(output_size, activation='softmax') # output layer\n",
    "])\n",
    "\n",
    "\n",
    "### Choose the optimizer and the loss function\n",
    "\n",
    "# we define the optimizer we'd like to use, \n",
    "# the loss function, \n",
    "# and the metrics we are interested in obtaining at each iteration\n",
    "model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy'])\n",
    "\n",
    "### Training\n",
    "# That's where we train the model we have built.\n",
    "\n",
    "# set the batch size\n",
    "batch_size = 100\n",
    "\n",
    "# set a maximum number of training epochs\n",
    "max_epochs = 100\n",
    "\n",
    "# set an early stopping mechanism\n",
    "# let's set patience=2, to be a bit tolerant against random validation loss increases\n",
    "early_stopping = tf.keras.callbacks.EarlyStopping(patience=2)\n",
    "\n",
    "# fit the model\n",
    "# note that this time the train, validation and test data are not iterable\n",
    "history = model.fit(train_inputs, # train inputs\n",
    "          train_targets, # train targets\n",
    "          batch_size=batch_size, # batch size\n",
    "          epochs=max_epochs, # epochs that we will train for (assuming early stopping doesn't kick in)\n",
    "          # callbacks are functions called by a task when a task is completed\n",
    "          # task here is to check if val_loss is increasing\n",
    "          callbacks=[early_stopping], # early stopping\n",
    "          validation_data=(validation_inputs, validation_targets), # validation data\n",
    "          verbose = 2 # making sure we get enough information about the training process\n",
    "          )  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA2oAAAI/CAYAAAAGHyr7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOzdd1hW9f/H8edhKyIi4gLEvQUU3DtTc6Tl3lvTsr37tXff6luZK8090zTTslw5ypGCiltx7wEq4EDW+f1xm18rB+gN5wZej+viuoT7nPN5kf/08pzz/himaSIiIiIiIiKOw8nqACIiIiIiIvJ3KmoiIiIiIiIORkVNRERERETEwaioiYiIiIiIOBgVNREREREREQejoiYiIiIiIuJgXKxauFChQmbJkiWtWl5ERERERMRSkZGRMaZp+t3qM8uKWsmSJYmIiLBqeREREREREUsZhnHkdp/p0UcREREREREHo6ImIiIiIiLiYFTUREREREREHIxl76iJiIiIiEjGJScnc/z4cRITE62OIunk4eFBQEAArq6u6T5HRU1EREREJBs5fvw4Xl5elCxZEsMwrI4jd2GaJrGxsRw/fpxSpUql+zw9+igiIiIiko0kJibi6+urkpZNGIaBr69vhu+AqqiJiIiIiGQzKmnZy738famoiYiIiIhIusTGxhIaGkpoaChFixbF39//xvdJSUl3PDciIoKnnnoqQ+uVLFmSmJiY+4mcbaXrHTXDMB4CvgKcgW9N0/z4H5+/CPS86ZqVAD/TNM/bMauIiIiIiFjI19eXrVu3AvD222+TL18+XnjhhRufp6Sk4OJy64oRHh5OeHh4luTMCe56R80wDGdgFNAKqAx0Nwyj8s3HmKb5qWmaoaZphgKvAqtV0kREREREcr5+/frx3HPP0bRpU15++WU2btxIvXr1qF69OvXq1WPv3r0ArFq1irZt2wK2kjdgwACaNGlC6dKlGTFiRLrXO3LkCM2aNSM4OJhmzZpx9OhRAObOnUvVqlUJCQmhUaNGAOzcuZNatWoRGhpKcHAw0dHRdv7tM0967qjVAvabpnkQwDCM2UB7YNdtju8OzLJPPBERERERcXT79u1j+fLlODs7Ex8fz5o1a3BxcWH58uW89tprzJs371/n7Nmzh5UrV5KQkECFChUYNmxYusbXDx8+nD59+tC3b18mTpzIU089xYIFC3j33XdZsmQJ/v7+XLx4EYCxY8fy9NNP07NnT5KSkkhNTbX7755Z0lPU/IFjN31/HKh9qwMNw8gLPAQMv/9oIiIiIiJyJ+8s2smuk/F2vWbl4vl56+EqGTqnc+fOODs7AxAXF0ffvn2Jjo7GMAySk5NveU6bNm1wd3fH3d2dwoULc+bMGQICAu661vr165k/fz4AvXv35qWXXgKgfv369OvXjy5dutChQwcA6tatywcffMDx48fp0KED5cqVy9DvZaX0DBO51YgS8zbHPgysvd1jj4ZhDDEMI8IwjIhz586lN6OIiIiIiDgwT0/PG39+4403aNq0KTt27GDRokW3HUvv7u5+48/Ozs6kpKTc09p/TVQcO3Ys77//PseOHSM0NJTY2Fh69OjBwoULyZMnDy1btuS33367pzWskJ47aseBwJu+DwBO3ubYbtzhsUfTNMcB4wDCw8NvV/ZERERERCQdMnrnKyvExcXh7+8PwOTJk+1+/Xr16jF79mx69+7NjBkzaNCgAQAHDhygdu3a1K5dm0WLFnHs2DHi4uIoXbo0Tz31FAcPHmTbtm088MADds+UGdJzR20TUM4wjFKGYbhhK2ML/3mQYRjeQGPgR/tGFBERERGR7OKll17i1VdfpX79+nZ5Jyw4OJiAgAACAgJ47rnnGDFiBJMmTSI4OJhp06bx1VdfAfDiiy9SrVo1qlatSqNGjQgJCeG7776jatWqhIaGsmfPHvr06XPfebKKYZp3v7FlGEZr4Ets4/knmqb5gWEYQwFM0xx7/Zh+wEOmaXZLz8Lh4eFmRETEveYWEREREcmVdu/eTaVKlayOIRl0q783wzAiTdO85Z4F6dpHzTTNxcDif/xs7D++nwxMzkBWERERERERuYX0PPooIiIiIiIiWUhFTURERERExMGoqImIiIiIiDgYFTUREREREREHo6ImIiIiIiLiYFTUbjJl7SFaffU7T87awlfLo/l52yn2nk7gWsr97/8gIiIiIpITNGnShCVLlvztZ19++SWPP/74Hc/5a2uu1q1bc/HixX8d8/bbb/PZZ5/dce0FCxawa9euG9+/+eabLF++PCPxb2nVqlW0bdv2vq9jT+kaz59b1Dw1nTbxUzkYV4zdu4rwZ1oxZpnFOUxx3HwCKF04P2UL57vxVcbPEy8PV6tji4iIiIhkme7duzN79mxatmx542ezZ8/m008/Tdf5ixcvvvtBt7FgwQLatm1L5cqVAXj33Xfv+VqOTkXtJpWrhYNxlEKx0dSMWYeRdOnGZ9eueHD8SDH2HCjC/rTirEkrxkGzGFfylaJYkUKU9bte3q6XOL987hiGYeFvIyIiIiJif506deL111/n2rVruLu7c/jwYU6ePEmDBg0YNmwYmzZt4urVq3Tq1Il33nnnX+eXLFmSiIgIChUqxAcffMDUqVMJDAzEz8+PsLAwAMaPH8+4ceNISkqibNmyTJs2ja1bt7Jw4UJWr17N+++/z7x583jvvfdo27YtnTp1YsWKFbzwwgukpKRQs2ZNxowZg7u7OyVLlqRv374sWrSI5ORk5s6dS8WKFdP1u86aNYsPP/wQ0zRp06YNn3zyCampqQwcOJCIiAgMw2DAgAE8++yzjBgxgrFjx+Li4kLlypWZPXv2ff13VlG7WYVWti/AME1IOA2x0RATjXvsfsrERFM6NhoubsQw02znJEPsSV/2HytKdGpRlpvFGWcW46xbCfIXLU1lfx+q+uenqr83pQt54uKsp01FREREJPvy9fWlVq1a/Prrr7Rv357Zs2fTtWtXDMPggw8+oGDBgqSmptKsWTO2bdtGcHDwLa8TGRnJ7Nmz2bJlCykpKdSoUeNGUevQoQODBw8G4PXXX2fChAk8+eSTtGvX7kYxu1liYiL9+vVjxYoVlC9fnj59+jBmzBieeeYZAAoVKsTmzZsZPXo0n332Gd9+++1df8+TJ0/y8ssvExkZiY+PDy1atGDBggUEBgZy4sQJduzYAXDjMc6PP/6YQ4cO4e7ufstHOzNKRe12DAPyF7N9lWr0vx8DpFyD8wchJhpio/GN2U/B2GhqnovA6VrcjWMTT7sTfcqfPX8GMMcM5JBTCczClSkeUJKq/gWo6u9NuSL5cHdxzvrfT0RERESyv19egdPb7XvNotWg1cd3POSvxx//KmoTJ04EYM6cOYwbN46UlBROnTrFrl27blvUfv/9dx599FHy5s0LQLt27W58tmPHDl5//XUuXrzIpUuX/vaY5a3s3buXUqVKUb58eQD69u3LqFGjbhS1Dh06ABAWFsb8+fPT8R8BNm3aRJMmTfDz8wOgZ8+erFmzhjfeeIODBw/y5JNP0qZNG1q0aAFAcHAwPXv25JFHHuGRRx5J1xp3oqJ2L1zcoXAl29d1Btfvwl2OuXEXzuPcHqqc2UWlM7twubLGdmAsxMV4smdLIFvSAphrBHK1QAXy+FelTFAgVYp7U6mYF3nd9FcjIiIiIo7pkUce4bnnnmPz5s1cvXqVGjVqcOjQIT777DM2bdqEj48P/fr1IzEx8Y7Xud2rQv369WPBggWEhIQwefJkVq1adcfrmKZ5x8/d3d0BcHZ2JiUl5Y7H3u2aPj4+REVFsWTJEkaNGsWcOXOYOHEiP//8M2vWrGHhwoW899577Ny5ExeXe/9/erUBezIMyOdn+wqqB9jGajoBXI6Fc7vhzC7yn91FyMkd1Ij5E9fk5ZAA7IHTu33YlxbATDOQi/nKYhSpjG+pYCqVKEpIYAE8XHXnTURERERucpc7X5klX758NGnShAEDBtC9e3cA4uPj8fT0xNvbmzNnzvDLL7/QpEmT216jUaNG9OvXj1deeYWUlBQWLVrEY489BkBCQgLFihUjOTmZGTNm4O/vD4CXlxcJCQn/ulbFihU5fPgw+/fvv/FOW+PGje/rd6xduzZPP/00MTEx+Pj4MGvWLJ588kliYmJwc3OjY8eOlClThn79+pGWlsaxY8do2rQpDRo0YObMmVy6dIkCBQrc8/oqalnF0xc8G0DJBhiAB4BpQvxJOLsb8+wu8h/fRujpXdSLW4FL4mI4AmmHDY6ahVlFEPEFKuMZVJ1SVetSoWw5nPW+m4iIiIhYpHv37nTo0OHG0IyQkBCqV69OlSpVKF26NPXr17/j+TVq1KBr166EhoYSFBREw4YNb3z23nvvUbt2bYKCgqhWrdqNctatWzcGDx7MiBEj+P77728c7+HhwaRJk+jcufONYSJDhw7N0O+zYsUKAgICbnw/d+5cPvroI5o2bYppmrRu3Zr27dsTFRVF//79SUuzzaz46KOPSE1NpVevXsTFxWGaJs8+++x9lTQA4263CTNLeHi4+ddeCvIPaalw4TCc3cXl49u5dCQK13PbKXjtxI1DYvHmdJ7ymEWrUqhcTYqUr4VRsAw4qbyJiIiI5GS7d++mUqVKdz9QHMqt/t4Mw4g0TTP8VsfrjpojcnIG3zLgWwbPSg/j+dfPE+O5eGgLx3atJ/FYFN5xuyl5cCpuhybBUrhm5CGhQAXylKiBZ4lQKBYMhSvb3qkTEREREZFsQ0UtO/HIT4FKjSlQyfa8rWmaHD13kZ1Rm4iN3oTL2e2Ujj1E5fOzIMo2eSfNcMEsVB7n4iFQNBhK1IZi1XXnTURERETEgamoZWOGYRBU2Ieg5i2geQvS0kx2n45ndvQ5ovdu59qxKMqZB6l6+jDVYpfgGzULANOzMEa5FlC+BZRuCh75Lf5NRERERETkZipqOYiTk0GV4t5UKe4NjcuSlNKerccusnZ/DKMOxHDs6GHqsJ02TttosP1H8mydjunkihFUD8o/BOVb2h65FBERERGHZprmbUfbi+O5l7kgGiaSi1y6lsLyXWdYGHWStftOE2LupUO+HTR3jcL3ykHbQb5loVxLW2krURdc3KwNLSIiIiJ/c+jQIby8vPD19VVZywZM0yQ2NpaEhARKlSr1t8/uNExERS2XOn85iV92nGLh1pNsPHwef87Sy2cPbTy2ERAXgZGaBG5eUKap7W5bueaQr7DVsUVERERyveTkZI4fP37XzaTFcXh4eBAQEICrq+vffq6iJnd0Ku4qP0WdYmHUSbafiCOvkUi/okd4NN8OylxYi9Ol04AB/jX+d7etWIhtg28REREREbknKmqSbgfPXWJR1Cl+jDrBwXOXcXGCXkFxdPHeTYX4tTif3AyY4FUMqjwKwV2gWKhKm4iIiIhIBqmoSYaZpsnOk/EsijrJwqiTnIpLxMPViUfKudHLdx+V4n/Hef8ySE0C33IQ3BWqdYKCpe5+cRERERERUVGT+5OWZhJx5AILo06wePtpzl9OwjuPK0Nq+tDfZxt598yDI2ttBwfWtt1lq9IB8ha0NriIiIiIiANTURO7SU5N44/9Mcz68yhLd53By92FfvVLMqiaC977f4Rtc+DcbnBygbLNIbgzVGgNrnmsji4iIiIi4lBU1CRT7D4Vz4gV0fyy4zT53F3oWy+IQfVL4XNpH2z7DrZ/DwmnbNMjK7eDap2hVCNwcrY6uoiIiIiI5VTUJFPtOR3P1yv2s3jHKfK6OtOnXkkGNyxNwTzOcPgP2D4Hdi2Ea/GQr6jtXbbgLlA0WENIRERERCTXUlGTLLHvTAIjVkTz8/ZT5HF1pnfdIIY0LI1vPndIvgr7ltgejYxeCmnJUKiCrbCFdAPvAKvji4iIiIhkKRU1yVLRZxIYuXI/i6JO4u5iK2yDG5bGz8vddsCV87Brga20HV0PTq4Q2gMaPgc+JS3NLiIiIiKSVVTUxBIHzl1i5G/7+XHrCdxcnOhVO4ghjUtT2MvjfwddOAzrRsLmKWCm2e6uNXweCpa2LLeIiIiISFZQURNLHTx3iZEr97NgywlcnZ3oWTuIoY1LUzj/TYUt/iT88SVEToa0FNu+bI1eAN8yluUWEREREclMKmriEA7HXGbkyv38sOUELk4G3WuVYFiTMhS5ubAlnIa1X0HERNtm2tU6Q6MXoVA564KLiIiIiGQCFTVxKEdiLzNq5X7mbz6Bk5NB37pBPPNgeTzdXf53UMIZWDcCNk2AlESo2hEavwR+FawLLiIiIiJiRypq4pCOnb/C179FMyfiOAE+efioQzUalvP7+0GXzsH6r2Hjt5B8Bao8arvDVqSyNaFFREREROxERU0c2qbD53l53jYOnrtMp7AAXm9TiQJ53f5+0OXY64VtPCRdgsrtodFLULSqNaFFRERERO6Tipo4vMTkVL7+LZqxqw/ik9eNd9tXoVXVohj/3BD7ynlYPwr+/AaSEqBiW2j8MhQLtia4iIiIiMg9UlGTbGPnyThenreNHSfiaVG5CO89UvXvw0b+cvUCbBgDG8bCtTio0Nr2Dlvx6lkfWkRERETkHqioSbaSkprGhD8O8d9l+3BzceL/Wleia83Af99dA7h60XZ3bcMoSIyDkO7Q4gPw9M364CIiIiIiGaCiJtnSoZjLvDJvG38eOk+9Mr581KEaQb6etz44MR7++MI2KdI9P7T8wFbablXuREREREQcgIqaZFtpaSazNx3jo8W7SU5L4/nmFehfvyQuzk63PuHMLvjpGTj2J5RqBG2/1KbZIiIiIuKQ7lTUbvN/uyKOwcnJoEftEix7rjENyvrxweLddBizjt2n4m99QpHK0P9XaPNfOBkFo+vC6k8hJSlrg4uIiIiI3AcVNckWinp7ML5PGCN7VOfEhas8/PUffL50L9dSUv99sJMT1BwIwzdChVaw8n34piEc3ZD1wUVERERE7oGKmmQbhmHQNrg4y59rTLvQ4nz9235af/U7EYfP3/oEr6LQZQr0mANJl2FiS1j0jG0AiYiIiIiIA1NRk2zHx9ON/3YJZXL/miQmp9H5m/W89eMOLl1LufUJ5VvC4xug7nDYPAVG1YId88Gi9zNFRERERO5GRU2yrSYVCrPk2Ub0rVuSqRuO0PKLNWw8dJu7a+75bJMgB68Er2LwfX+Y2QUuHs3a0CIiIiIi6aCiJtlaPncX3m5Xhe+H1sXNxYnu4zfw7e8Hue000+KhMGgFtPwIDq+FUbVh3deQepu7cSIiIiIiFlBRkxwhLKggPw6vz4OVCvP+z7sZPnPL7R+FdHaBuo/DE39Cqcaw9HUY3xRObM7a0CIiIiIit6GiJjlGfg9XxvYK45VWFfllxynaj/yD/WcTbn9CgUDoPgu6TIVLZ+HbZvDLK3DtDueIiIiIiGQBFTXJUQzDYGjjMswYVIe4q8m0G7mWn7advNMJULm9bZR/+AD4cyyMqgN7f8260CIiIiIi/6CiJjlS3TK+/PRkQyoVy8/wmVt4Z9FOklPTbn+Chze0+RwGLrX9eVZX+PVVbZQtIiIiIpZQUZMcq6i3B7MG16FfvZJMWnuY7uM2cCY+8c4nBdaCIaug9lDYMBomPaTJkCIiIiKS5VTUJEdzc3Hi7XZVGNG9OrtOxdNmxB9sOBh755Nc3KDVJ7Z312KiYWxD2PtL1gQWEREREUFFTXKJdiHFWfBEffLncaHnt38ybs2B24/w/0vl9vDYavAJglndbNMhU5OzJrCIiIiI5GoqapJrlC/ixY9P1KdF5SJ8uHgPj8/YTELiXYpXwdIwYCnUHGTbb21yG4g7njWBRURERCTXUlGTXMXLw5XRPWvwf60rsXTXGdqPXMu+M3cZx+/qYRs00mkinNllexRy39KsCSwiIiIiuZKKmuQ6hmEwuFFpZg6qTXxiCu1HruXHrSfufmLVjrZBI/mLw8zOsPxtSL3NptoiIiIiIvdBRU1yrdqlffn5qQZU9c/P07O38vbCnSSl3GGEP0ChsjBoOYT1gz++gCkPQ/wd9mkTEREREbkHKmqSqxXJ78HMwXUY2KAUk9cdptu49ZyKu3rnk1zzwMNfQYfxcCoKxjaA/SuyJrCIiIiI5AoqapLruTo78UbbyozsUZ09pxNoO+IP1h2IufuJwV1sj0LmKwLTO8Jv70NaambHFREREZFcQEVN5Lq2wcVZOLw+BfK60mfCRuZGHLv7SX7lYdAKCO0Jaz6Fqe0h4XTmhxURERGRHE1FTeQmZQt7seCJ+tQp7cuL32/j6xXRd99vzS0vPDIKHhkDxyNsUyEPrsqSvCIiIiKSM6moifyDl4crE/vVpEN1fz5fto/XfthBSupdhowAhPaAISshjw9MfQRWfaxHIUVERETknqioidyCm4sTn3cJ4YmmZZi18ShDpkVyJSkdo/gLV4LBv0FwV1j1EUzvAJfOZX5gEREREclRVNREbsMwDF5sWZH3H6nKqr1n6T5uAzGXrt39RPd88OhYaPc1HN0A3zSyPRIpIiIiIpJOKmoid9GrThDf9A5n75kEOo5Zx+GYy3c/yTCgRh8YuBScXWBSK4icnOlZRURERCRnUFETSYfmlYswc3AdEhJT6DBmHVuOXkjficVCYMhqKNkAFj0NC5+ElHTclRMRERGRXE1FTSSdapTwYd6wenh5uNB9/AaW7TqTvhPzFoSe30PD52HzVNvdtbjjmRtWRERERLI1FTWRDChVyJN5w+pRoYgXj02LYPqGI+k70ckZmr0JXafDuX3wTWM49HvmhhURERGRbEtFTSSDCuVzZ9aQOjSpUJjXF+zg0yV77r7X2l8qPWybCpm3oG1z7HUjIb3nioiIiEiuoaImcg/yurkwrncY3WsFMmrlAZ6fG0VSSjr2WgPwKw+DVkCFVrD0/2DeQEhKx4ASEREREck1VNRE7pGLsxMfPlqN55uXZ/7mEwycsomExOT0neyR3/YYZLM3Ycd8+PZBiD2QuYFFREREJNtQURO5D4Zh8GSzcnzaKZj1B2Lp+s0GzsQnpvdk24CRXvMg4RSMawr7lmRuYBERERHJFlTUROygc3ggE/rV5EjsZTqMXkf0mYT0n1y2GQxZBT4lYGZXWPUJpKXzMUoRERERyZFU1ETspHF5P757rC5JqWl0HLOOjYfOp/9kn5IwYCkEd4VVH8Ls7nD1YqZlFRERERHHpqImYkdV/b2ZP6wehbzc6TXhTxZvP5X+k93ywqNjodWnsH85jH8AzuzKvLAiIiIi4rBU1ETsLLBgXuYPq0ewvzdPzNzM7I1H03+yYUDtIdD3J0i6ZBsysmN+5oUVEREREYekoiaSCQrkdWP6oNo0Lu/Haz9sz9idNYCgujBkNRStCt/3h6WvQ2pK5oQVEREREYejoiaSSTxcnRnTM4waJXx4evYWfo8+l7EL5C9mu7NWcxCs+xqmPwpXMvDem4iIiIhkWypqIpkoj5szE/rVpIxfPh6bFsmWoxcydgEXN2jzObQfDUc3wITm2m9NREREJBdQURPJZN55XJk6sBZ+Xu70n7wpY6P7/1K9J/T50XZH7dtmcGSd/YOKiIiIiMNQURPJAoW9PJg2oDZuzk70nrCR4xeuZPwiQfVg0HLI6wtT20PUd/YPKiIiIiIOQUVNJIuU8M3L1IG1uJKUQu8JG4m5dC3jF/EtAwOXQWBt+GEIrPwQTNP+YUVERETEUipqIlmoYtH8TOpfk1NxV+k7cSPxickZv0jegtBrPoT2hNWfwLxBkJxo/7AiIiIiYhkVNZEsFhZUkLG9wth7OoFBUyJITE7N+EVc3KD9KGj2Juz43vYo5OUY+4cVEREREUukq6gZhvGQYRh7DcPYbxjGK7c5polhGFsNw9hpGMZq+8YUyVmaVCjMf7uGsunweYbP3EJKalrGL2IY0PB56DwZTm21DRk5t8/uWUVEREQk6921qBmG4QyMAloBlYHuhmFU/scxBYDRQDvTNKsAnTMhq0iO0i6kOO+2r8ry3Wd4ad420tLu8V2zKo/a9ltLugwTHoSD+ncSERERkewuPXfUagH7TdM8aJpmEjAbaP+PY3oA803TPApgmuZZ+8YUyZl61wniueblmb/5BO//vBvzXgeDBNaEQSvAqxhM7wCbp9k3qIiIiIhkqfQUNX/g2E3fH7/+s5uVB3wMw1hlGEakYRh97BVQJKd78oGy9K9fkolrDzFq5f57v5BPEAxcCiUbwsLhsPxtSLuHRypFRERExHIu6TjGuMXP/vnP/i5AGNAMyAOsNwxjg2maf3thxjCMIcAQgBIlSmQ8rUgOZBgGb7SpzMUryXy2dB8F8rrRq07QvV3Mwxt6zoXFL8IfX8D5g/DoN+Cax76hRURERCRTpaeoHQcCb/o+ADh5i2NiTNO8DFw2DGMNEAL8raiZpjkOGAcQHh6uzZ9ErnNyMvhPp2Dirybzxo878M7jysMhxe/tYs6u0PYL8C0LS1+HuOPQfTbkK2zf0CIiIiKSadLz6OMmoJxhGKUMw3ADugEL/3HMj0BDwzBcDMPIC9QGdts3qkjO5ursxKieNagZVJDn5mxl9b5z934xw4B6w6HrdDi7G8Y3gzO77BdWRERERDLVXYuaaZopwHBgCbbyNcc0zZ2GYQw1DGPo9WN2A78C24CNwLemae7IvNgiOZOHqzPf9gunbGEvhk6LJPLIhfu7YKW20H8xpCbBxJawf7l9goqIiIhIpjLuecrcfQoPDzcjIiIsWVvE0Z1LuEbnseu4cCWZOY/VpUJRr/u7YNwJmNkVzu6C1p9CzYH2CSoiIiIi98wwjEjTNMNv9Vm6NrwWkazl5+XOtIG18XB1oveEPzl2/sr9XdDbHwb8AmUfhJ+fs727pomQIiIiIg5LRU3EQQUWzMvUAbW5lpJGrwl/ci7h2v1d0N0Lus+CmoNh3dewYBikJtsnrIiIiIjYlYqaiAOrUNSLSf1rcjb+Gn0mbiQ+8T6LlZOz7dHHB16HbbNhVndIumyfsCIiIiJiNypqIg6uRgkfxvYOI/pMAo9P30xSyn0+smgY0OhFePgrOLACpjwMl2PtE1ZERERE7EJFTSQbaFzej487BvPH/hhembcNu6y1GxQAACAASURBVAwBCusHXabBmZ22iZAXj97/NUVERETELlTURLKJTmEBPN+8PPO3nODzpfvufkJ6VGoLvX+AS2dhQgvttSYiIiLiIFTURLKR4Q+UpXutQEau3M+MP4/Y56JB9WwTIQEmPQRH1tvnuiIiIiJyz1TURLIRwzB4r31Vmlbw440FO1ix+4x9LlykCgxcCp6FYdojsGexfa4rIiIiIvdERU0km3FxdmJkjxpU9fdm+MwtbD120T4XLlACBiyxlbbvekLkFPtcV0REREQyTEVNJBvydHdhQt+aFPJyY+DkTRyJtdOIfU9f6LMQSjeFRU/Bms/AHoNLRERERCRDVNREsik/L3em9K9FmmnSd+JGYi/d54bYf3HPBz2+g+Cu8Nt78MvLkHafWwKIiIiISIaoqIlkY6X98vFt33BOxSUyaGoEV5NS7XNhZ1d4ZCzUHQ4bv4F5AyDFTkVQRERERO5KRU0kmwsLKshX3aqz9dhFnpq9hdQ0Oz2q6OQELT+A5u/Bzh9gRmdIjLfPtUVERETkjlTURHKAh6oW5a22lVm26wzvLNppnw2x/1L/KdvdtcN/wJS2tj3XRERERCRTqaiJ5BD96pdiSKPSTF1/hHFrDtr34qHdoftsiIm2bYx9/pB9ry8iIiIif6OiJpKDvPJQRdoGF+OjX/bw49YT9r14+Ra2iZCJF21l7VSUfa8vIiIiIjeoqInkIE5OBp93CaF2qYK8MDeKdQdi7LtAYE3bXmvObjCpDRxaY9/ri4iIiAigoiaS47i7ODOudzglfT15bFoke08n2HcBvwowcCl4B8D0jrD9e/teX0RERERU1ERyIu+8rkweUIs8rs70m7SR03GJdl7AHwb8Av7hMG8grB2hjbFFRERE7EhFTSSH8i+Qh0n9a5KQmEK/SRuJT0y27wJ5fKD3D1D5EVj2xvWNse20j5uIiIhILqeiJpKDVSnuzZheNdh/9hLDpkeSlJJm3wVcPaDTpP9tjD2nDyRfte8aIiIiIrmQippIDtewnB+fdAxm7f5YXp63zb57rMH/NsZ+6GPY8zNMaQeXY+27hoiIiEguo6Imkgt0DAvghRbl+WHLCT5bujdzFqkzDLpMsY3tn6i91kRERETuh4qaSC7xRNOydK9VglErDzBr49HMWaRye+i7EK7EwoTmcCIyc9YRERERyeFU1ERyCcMweK99FRqX9+PNH3cQeeRC5ixUog4MWAqueWByW9i3JHPWEREREcnBVNREchEXZye+6hZKMe88PD4jkrMJdh7b/xe/8jBwORQqB7O6QcSkzFlHREREJIdSURPJZQrkdWNsrzDiriYzfMYWklPtPAnyL15FoN9iKNMMfnoGfntfe62JiIiIpJOKmkguVLl4fj7pGMzGw+f54OfdmbeQez7oPhtq9IE1n8KCYZCSlHnriYiIiOQQLlYHEBFrtA/1J+pYHBPXHiIk0JtHqwdkzkLOLvDwCPAOhJUfQMIp6DINPPJnznoiIiIiOYDuqInkYq+2rkjtUgV5df52dp6My7yFDAMavwTtR8PhP2BSK4g/mXnriYiIiGRzKmoiuZirsxMje9SgQB43HpsWycUrmfxYYvWe0OM7uHAYvm0OZzPxsUsRERGRbExFTSSX8/NyZ0yvGpyNv8aTs7aQmpbJAz/KPgj9F0NaCkxoCYd+z9z1RERERLIhFTURoXoJH95pX4Xfo2P477K9mb9gsRAYtAy8isL0DrD9+8xfU0RERCQbUVETEQC61ypBt5qBjFp5gF93nM78BQuUgIFLIKAmzBsIv/9X4/tFRERErlNRE5Eb3mlfhZDAArwwN4r9Zy9l/oJ5fKDXfKjaEVa8A/OHQPLVzF9XRERExMGpqInIDe4uzozpWQN3FycemxZBQmJy5i/q6gEdJ8ADr8P2OZoIKSIiIoKKmoj8Q/ECeRjZowaHY6/wwtwo0jJ7uAjYxvc3ehG6zYSYaBjXBI5HZP66IiIiIg5KRU1E/qVuGV9ebVWRJTvPMGb1gaxbuGIbGLgMXDxgUmvYOivr1hYRERFxICpqInJLAxuU4uGQ4ny2dC+r953LuoWLVIYhqyCwFiwYCkv+D1JTsm59EREREQegoiYit2QYBp90rEaFIl48NWsLx85fybrF8xaE3j9ArSGwfiTM7AJXL2bd+iIiIiIWU1ETkdvK6+bCN73DME2Tx6ZFcjUpNesWd3aF1p/Cw1/BoTXwbTPb+2siIiIiuYCKmojcUZCvJ191q87u0/G89sN2zKze6yysH/RdaLujNr4ZRC/L2vVFRERELKCiJiJ31bRiYZ59sDw/bDnBlHWHsz5AUD0YshJ8Stgeg1w7Qptji4iISI6moiYi6TK8aVkerFSE93/ezcZD57M+QIESMGAJVGoHy96AHx6D5MSszyEiIiKSBVTURCRdnJwM/ts1hMCCeXl8xmZOx1lQktw8ofNkaPo6bPsOJreG+FNZn0NEREQkk6moiUi65fdw5ZveYVxJSmHYjEiupWThcJG/GAY0fhG6Toeze65vjh2Z9TlEREREMpGKmohkSPkiXnzWOYQtRy/yzqJd1gWp9DAMWgYu7jCpFUTNti6LiIiIiJ2pqIlIhrWuVozHGpdm5p9HmRNxzLogRarA4JW2zbF/eAyWvgFpFtzlExEREbEzFTURuScvtqhAvTK+vL5gB9uPx1kXxNPXtjl2zUGwbgTM7AqJFuYRERERsQMVNRG5Jy7OTnzdvTqFPN0YOj2SC5eTrAvj7AptPoe2X8LBlbb91mIPWJdHRERE5D6pqInIPfPN586YXmGcS7jGU7O3kJpm8d5m4f2hz49wJRbGN4UDv1mbR0REROQeqaiJyH0JCSzAO+2r8Ht0DF8s22d1HCjZwLY5dv4AmN4JNozV5tgiIiKS7aioich961YzkC7hAYxcuZ+lO09bHQd8SsLAJVD+Ifj1ZVj4JKRY+GimiIiISAapqInIfTMMg3fbV6WavzfPz4ni4LlLVkcCdy/bXmsNX4At02BqO7h0zupUIiIiIumioiYiduHh6syYXjVwcTYYOj2Sy9dSrI4ETk7Q7A3oOAFOboHxD8Dp7VanEhEREbkrFTURsZsAn7yM6F6d/Wcv8fK8bZiO8m5YtU7Q/xdIS4EJLWH3IqsTiYiIiNyRipqI2FXDcn4836ICP207xcS1h62O8z/+NWxDRgpXgu96wer/aMiIiIiIOCwVNRGxu8eblKFF5SJ8uHg3fx6MtTrO/3gVhX4/Q3A3WPkBzO0HSVesTiUiIiLyLypqImJ3hmHwWZcQggrm5YmZWzgTn2h1pP9x9YBHx0Lzd2HXjzCxJcQdtzqViIiIyN+oqIlIpsjv4crY3mFcSUph2PRIklLSrI70P4YB9Z+GHt/B+UMwrikc22h1KhEREZEbVNREJNOUL+LFfzoFs/noRT74eZfVcf6tfEsYtBzcPGFyG9g60+pEIiIiIoCKmohksrbBxRnUoBRT1h/hhy0O+Ihh4Yow+DcoUQcWDIMl/wdpqVanEhERkVxORU1EMt0rrSpSu1RBXp2/nV0n462O8295C0Kv+VBzMKwfCTO7QGKc1alEREQkF1NRE5FM5+LsxMgeNfDO48rQ6ZHEXUm2OtK/ObtCm8+g7RdwcBWMbwaxB6xOJSIiIrmUipqIZAk/L3dG9wzjVNxVnvluC2lpDrqHWfgA6PMjXImF8Q/AgZVWJxIREZFcSEVNRLJMWJAPb7atzMq95xjxW7TVcW6vZAPb5tj5i8P0jvDnOG2OLSIiIllKRU1EslSvOkF0qOHPVyuiWbnnrNVxbs+nJAxcCuVawC8vwk/PQqoDPrIpIiIiOZKKmohkKcMw+PDRalQqmp+nZ2/hSOxlqyPdnrsXdJsJDZ6FyEkw9RG4HGt1KhEREckFVNREJMt5uDrzTe8wDMNg6PTNXE1y4HH4Tk7w4Nvw6Dg4vgnGN4Wzu61OJSIiIjmcipqIWCKwYF6+7BbKntPxvPbDdkxHfwcspCv0XwwpifDtg7D3V6sTiYiISA6moiYilmlaoTDPNCvPD1tOMHX9Eavj3F1AOAxeCb5lYVY3+ONLDRkRERGRTKGiJiKWevKBsjSrWJj3f97FlqMXrI5zd97+0P8XqPIoLH8LfhgKyYlWpxIREZEcRkVNRCzl5GTweZcQCnt5MHzmFi5cTrI60t255YVOE6Hp67BtNkxuAwlnrE4lIiIiOYiKmohYrkBeN0b3rMG5hGs8N2er426GfTPDgMYvQpdpcHaXbcjIya1WpxIREZEcQkVNRBxCSGAB3mhbiZV7zzFm9QGr46Rf5XYwYAkYTjDxIdj5g9WJREREJAdQURMRh9GrThAPhxTn86V7WXcgxuo46Vcs2DZkpFgIzO0HKz+CtDSrU4mIiEg2pqImIg7DMAw+6lCNkoU8eWrWVs7GZ6MhHfn8oO9CCO0Fqz+GuX0hyYE38xYRERGHpqImIg4ln7sLY3qGcelaMsNnbSElNRvdmXJxh/YjocUHsOcnmNgSLh6zOpWIiIhkQypqIuJwKhT14sNHq7Hx0Hk+X7bP6jgZYxhQbzj0mAMXjsD4B+DYRqtTiYiISDajoiYiDqlDjQC61wpkzKoDrNidDUffl2sOg5aDez7b+P7N06xOJCIiItmIipqIOKy3Hq5C5WL5eW5OFMfOX7E6Tsb5VYBBKyCoPiwcDj89BynZYJ84ERERsZyKmog4LA9XZ8b0qkGaafLEzM1cS0m1OlLG5S0IveZB/WcgYgJMaQvxp6xOJSIiIg5ORU1EHFqQryefdgph2/E4Pvh5t9Vx7o2TMzR/BzpPhtM7YFxjOLrB6lQiIiLiwFTURMThPVS1KIMalGLq+iMsjDppdZx7V+VR23trbp4wuS1sHA+maXUqERERcUAqaiKSLbzcqiJhQT68Mm8b+89esjrOvStS2bY5dpmmsPgF+HE4JGej/eJEREQkS6SrqBmG8ZBhGHsNw9hvGMYrt/i8iWEYcYZhbL3+9ab9o4pIbubq7MTIHtXxcHXm8RmRXElKsTrSvctTALp/B41fhq3TYdJD2m9NRERE/uauRc0wDGdgFNAKqAx0Nwyj8i0O/d00zdDrX+/aOaeICMW88/BVt1Ciz17i9R92YGbnxwadnKDpa9BtFsTst723dmiN1alERETEQaTnjlotYL9pmgdN00wCZgPtMzeWiMitNSznx1MPlGP+lhPM3pQD7kJVbA1DVkJeX5j6CKwfpffWREREJF1FzR+4+f+Gjl//2T/VNQwjyjCMXwzDqGKXdCIit/BUs3I0LFeItxbuZMeJOKvj3L9C5WDwb7bStuQ1mDcIkrLhvnEiIiJiN+kpasYtfvbPf+7dDASZphkCfA0suOWFDGOIYRgRhmFEnDt3LmNJRUSuc3Yy+LJrKAXzuvHEzM3EXU22OtL9c/eCLtOg2ZuwYx5MaA7nD1mdSkRERCySnqJ2HAi86fsA4G/zsU3TjDdN89L1Py8GXA3DKPTPC5mmOc40zXDTNMP9/PzuI7aI5Ha++dwZ2aM6xy9c5cW5Udn7fbW/GAY0fB56fg9xx2FcE9i/3OpUIiIiYoH0FLVNQDnDMEoZhuEGdAMW3nyAYRhFDcMwrv+51vXrxto7rIjIzcJLFuTVVhVZuusME/7IQXefyj0IQ1aBdwBM7wS/f6731kRERHKZuxY10zRTgOHAEmA3MMc0zZ2GYQw1DGPo9cM6ATsMw4gCRgDdzBzxz9si4ugGNihFyypF+PiXPUQcPm91HPspWAoGLoWqHWDFuzCnN1xLsDqViIiIZBHDqj4VHh5uRkREWLK2iOQscVeTefjrP0hKSePnpxrgm8/d6kj2Y5q2SZDL3gTfstBthm34iIiIiGR7hmFEmqYZfqvP0rXhtYiII/PO48ronjU4fyWJZ77bSmpaDrqhbxhQbzj0WQBXYmzvrW2dpUchRUREcjgVNRHJEar6e/NOuyr8Hh3Df5fttTqO/ZVqBENWQ7EQWDAUvu8PVy9YnUpEREQyiYqaiOQY3WoG0q1mIKNWHmDx9lNWx7G/AoHQd5FthP/uRTCmPhz63epUIiIikglU1EQkxzAMg3faV6F6iQK8MDeKPafjrY5kf07OthH+A5eBiwdMeRiWvQUpSVYnExERETtSURORHMXdxZmxvcLwdHdhyNRILl7JoQXGvwYM/R1q9IG1X8KEB+HcPqtTiYiIiJ2oqIlIjlMkvwdje4VxKu4qT87aQkpqmtWRMoebJ7QbAV1nwMVj8E0j2DRBg0ZERERyABU1EcmRwoJ8eK99VX6PjuHTJTlwuMjNKrWFYesgqC78/BzM6g6XY6xOJSIiIvdBRU1EcqxutUrQq04JvllzkB+3nrA6TubKXwx6zoOWH8GBFTCmHkQvtzqViIiI3CMVNRHJ0d5sW4WaJX14ed42dpyIszpO5nJygrqPw+CVkKcgzOgIv7wMyYlWJxMREZEMUlETkRzNzcWJ0T3DKJDHjcemRXL+cg4dLnKzolVhyEqoPRT+HAvjm8KZnVanEhERkQxQURORHM/Py51veodx7tI1npixOecOF7mZax5o9YntccjLMTCuCawfDWm54HcXERHJAVTURCRXCAkswEePVmP9wVg+XLzH6jhZp9yD8Ph6KNMMlrwK0ztAfA7cDFxERCSHUVETkVyjY1gA/euXZOLaQ8yLPG51nKzjWQi6z4K2X8DRDbZBI7t/sjqViIiI3IGKmojkKq+1rkSd0gV59YftbDt+0eo4WccwIHwAPLYGCgTCdz1h4VOQdNnqZCIiInILKmoikqu4OjsxqkcN/PK589i0SM4lXLM6UtbyKw8Dl0P9Z2DzVBjbEI5HWp1KRERE/kFFTURyHd98tuEiF64k8fiMSJJSctmADRc3aP4O9F0EKddgQnNY/SmkplidTERERK5TURORXKmqvzefdAxm0+ELvPfTLqvjWKNUQxj2B1R5BFa+D5PbwIXDVqcSERERVNREJBdrH+rPkEalmbbhCN9tOmp1HGvk8YFOE6HDeDi7C8Y0gK2zwDStTiYiIpKrqaiJSK72UssKNCxXiDcW7GTz0QtWx7FOcBcYthaKVoMFQ+H7/nDlvNWpREREci0VNRHJ1Vycnfi6e3WKenswdFokZ+MTrY5knQIloN9P0Owt2L0IxtSHg6utTiUiIpIrqaiJSK5XIK8b4/qEkZCYwtDpkVxLSbU6knWcnKHhczBoObh5wtR2sOT/bENHREREJMuoqImIABWL5uezziFsPnqRt37ciZnb39EqXh0eW23be239SBj/AJzdbXUqERGRXENFTUTkujbBxXi8SRlmbzrGjD9z6XCRm7l5QtsvoPt3kHAavmkMG8Zq0IiIiEgWUFETEbnJ8y0q0KSCH28v3MnGQxqmAUCFh+Dx9VC6Cfz6MkzvaCtuIiIikmlU1EREbuLsZPBVt+oEFszL4zMiOXHxqtWRHEO+wtDjO2jzORxZB6Prwu6frE4lIiKSY6moiYj8g3ceV8b3CeNachqDp0RwJSnF6kiOwTCg5iDbu2sFAuG7nrDwSbh2yepkIiIiOY6KmojILZQt7MWI7tXZfTqeF+ZGabjIzfwqwMDl0OBZ2DwNxjaA6GVWpxIREclRVNRERG6jacXCvPJQRRZvP82IFfutjuNYXNzgwbdt+64ZTjCjE8zqARcOWxxMREQkZ1BRExG5gyGNStOhuj9fLN/HrztOWR3H8ZRsYBs08uDbcHAVjKoNKz+CZL3bJyIicj9U1ERE7sAwDD7sUI3QwAI8+10Uu07GWx3J8bi42x6DHL4JKraB1R/DqFq2YSN6ZFREROSeqKiJiNyFh6sz43qHkT+PC4OnRhBz6ZrVkRyTtz90mgh9fwJXT9uwkRmdIEaPjYqIiGSUipqISDoUzu/B+D7hxFy6xuPTN5OUkmZ1JMdVqiEM/R1afgTHNsLoOrD8bU2HFBERyQAVNRGRdAoOKMB/OgWz8fB53vxxhyZB3omzK9R9HIZHQLXO8McXMLIm7JinxyFFRETSQUVNRCQD2of680TTMszedIwp6w5bHcfxeRWBR8fAgKXgWQi+HwBTHoYzu6xOJiIi4tBU1EREMuj55hV4sFIR3vt5N39Ex1gdJ3soURuGrII2/4XT2217r/36GiTGWZ1MRETEIamoiYhkkJOTwZfdQinj58kTMzdzKOay1ZGyBydnqDkQntwMNfrAhtHwdThsnQVpeudPRETkZipqIiL3IJ+7C9/2qYmTAYOnRhCfmGx1pOzD0xce/hKGrIQCJWDBUJj0EJyKsjqZiIiIw1BRExG5RyV88zK6ZxiHYy7z9KwtpKZpSEaGFK8OA5dB+9EQewDGNYHFL0Ki9qoTERFRURMRuQ91y/jyVrsqrNx7jv8s2WN1nOzHyQmq94QnI6HmYNg43jbOf+8vVicTERGxlIqaiMh96l0niF51SvDN6oPM33zc6jjZU54C0Po/MGg5eBSAWd1gTl9IOG11MhEREUuoqImI2MFbD1ehTumCvDJ/O1uOXrA6TvYVEA6PrYYH3rDdVRtZCyIna9iIiIjkOipqIiJ24OrsxOieYRTJ785j0yI5HZdodaTsy9kVGr0Aw9ZBsWBY9DRMaQsx0VYnExERyTIqaiIidlLQ041v+9Tk8rUUhkyLIDE51epI2VuhstB3EbT7Gs7sgDH1YfWnkJJkdTIREZFMp6ImImJHFYp68WW36mw/EcdL32/DNDUJ8r4Yhm3PtSc2QcXWsPJ9GNcYjm2yOpmIiEimUlETEbGz5pWL8EKLCiyMOsnoVQesjpMzeBWBzpOh+2xIjIMJzW2j/K8lWJ1MREQkU6ioiYhkgseblKFdSHE+W7qXZbvOWB0n56jQCp74E2oNsY3yH1Vbo/xFRCRHUlETEckEhmHwn07BVC3uzTOzt7D3tO782I27l22U/8Bl4OF90yh/FWIREck5VNRERDKJh6sz4/uE4+nuwsApm4i5dM3qSDlLYE0YshoeeN12V21UTYic8v/s3XeUVeW9//H3nk6vQ+9l6B1pCqIUewV7jRp7oinGmF+Sm3JzTYzJ1eTaULEbk2BLTFQQaSJFioL03plCZ4bp+/fHmagxiiAz7DMz79das4Zz5uTMh7V2cD7zPPv7gPcFSpKqAIuaJFWgZvXSePzqgWQfKOCm5xY6CbK8JaXAiLtio/yb9oK/fxuePhty1kadTJKkY2JRk6QK1qd1fX5/cV8WbtrDPa8sdRJkRfjXKP9z/gCZS+GRYfDeA1BSHHUySZK+FouaJB0HZ/VuzvfGZPDq4m08NM3VngqRkAADromN8s8YC+/8F0wcC1kro04mSdJRs6hJ0nFy+6mdOL9vC+6fvJp/Lt0RdZyqq05TuPg5GD8Rdm+Ax4bDrN+7uiZJqlQsapJ0nARBwK/H9aZ/m/p89y8fsmTr3qgjVV1BAD3HwW3zYyP9p/4cnhwNmcujTiZJ0hGxqEnScZSWnMiEqwfSqFYqNzyzgB37DkUdqWqrnQ4XPxs7LHvvFphwMsz8LZQURZ1MkqTDsqhJ0nHWuHYqT147kNyCYm54ZgF5hW7Jq3A9LogdlN31LHj3v+GJUbDz46hTSZL0pSxqkhSBrs3q8sfL+7Fix37ufOlDSkudBFnhajWOraxd/Czs3w4TRsKM+1xdkyTFJYuaJEXk1K5N+X9ndWfy8kx+O3lV1HGqj+7nwa3zYp+n/QoePxV2Lo06lSRJ/8aiJkkRuu7Edlw+uA2PTF/HpIVbo45TfdRqBOOfhEuehwM7Y6tr038NxYVRJ5MkCbCoSVKkgiDg5+f24MROjbjnlSXM37A76kjVS7dzYveu9bgQpt8bW13bsSTqVJIkWdQkKWrJiQk8fPkAWjeoyU3PLWDTrtyoI1UvNRvCuMfh0hchNwsePwWm/Y+ra5KkSFnUJCkO1KuZzJPXnkBpCNc/s4D9+Q64OO66ngW3zoWe42HGb2LbIbd/GHUqSVI1ZVGTpDjRvnEtHr1yABtzcrnthUUUl5RGHan6qdkQLnwMLnsJ8nbFtkJO/jEc2hN1MklSNWNRk6Q4MrRjI/77/J7MWpPDL99YHnWc6qvLGXDbXOhzGbz/f/BgX5j9ByjKjzqZJKmasKhJUpy5dFAbvjm8Pc/M2cSzczZGHaf6qtEAzn8IbpoJLQfAlJ/AHwfAhy9CaUnU6SRJVZxFTZLi0A/P6Maork34+d+XM2N1dtRxqrfmveGqV+Dq12OHZr92Czw6HFZPhtCDyiVJFcOiJklxKDEh4MHL+tG5SW1uf2ERazIPRB1JHUbCN6fB+KegKA9evAiePhu2Log6mSSpCrKoSVKcqp2axBPXDCQ1OZHrn1nA7lzHxUcuIQF6Xgi3zYcz74ecVfDEKPjL1ZCzNup0kqQqxKImSXGsVYOaTLh6ADv353PzcwspKPbeqLiQlAKDvgnfXgwj74E178BDg+CN78CBnVGnkyRVARY1SYpz/ds04P6L+jB/427ueWUpofdFxY/UOjDyh3DHhzDwOlj0LPyhH7z7K8jfH3U6SVIlZlGTpErg3D4tuHN0Z15ZtI373l4VdRx9Xu0mcNb9sS2RGafDzPvgD31h7qNQ7JZVSdLRs6hJUiVxx6jOXD64DY9MX8cTs9ZHHUdfpFFHuOip2NCRpj3grbvh/wbC0klQ6gHmkqQjZ1GTpEoiCAJ+eV5PzuzVjP/+xwpeXrg16kj6Mi37w9V/gytfhtS68PL1MOFkR/pLko6YRU2SKpHEhID/vaQvwzo24gcvL2HqisyoI+nLBAF0Gh07MPvCxyF/b2yk/xOjYe1UC5sk6bAsapJUyaQmJTLh6oF0b16XW19YxAcbd0cdSYeTkAC9L4bbF8I5D8amQj5/IUw8HdbPiDqdJClOWdQkqRKqnZrE0984gZb1a3D90x+wcqcTBuNeUgoMuBa+vQjO+h3s3QzPnhs7NHvTcWUsLAAAIABJREFU+1GnkyTFGYuaJFVSjWqn8uz1g6iRksjVT85ny+68qCPpSCSlwgk3xM5gO+M+yFkNT50Bz54HW+ZHnU6SFCcsapJUibVqUJPnrh9MQXEpVz05j+wDBVFH0pFKToPBN8EdH8HYX8HOj+HJMfD8ONi6MOp0kqSIWdQkqZLLaFqHideewM79+Vz71Hz25xdFHUlHI7kGDLsd7lwCo38O2xbBE6fCi5fA9g+jTidJiohFTZKqgAFtG/DIlQNYtfMANz67gPyikqgj6Wil1IKT7owVtlN/Apvnxkb6v3QF7FwadTpJ0nFmUZOkKuKULk24/6I+zF2/mzteWkxxiQcsV0qpdWDE92OFbeSPYMNMePQk+MvVkLUi6nSSpOPEoiZJVcj5/Vry07O78/ayTH782seEntVVeaXVg5F3xwrbiLtiZ689PBQmXQdZK6NOJ0mqYElRB5Akla/rTmrP7txC/m/aWhrWSuEHp3eNOpKORY0GcOqPYcit8P4fYN4E+PhlaDccBn4Dup4TG/0vSapSLGqSVAV9b2wGu3ILeXj6OhrWSuGG4R2ijqRjVbMhjP4ZDL0dFj0LC5+Kra7VSod+V8GAa6BBu4hDSpLKSxDVtpiBAweGCxYsiOR7S1J1UFIactsLi3hr2U5+f3EfLuzfKupIKk+lpbDuXVgwEVa/CWEInUbDwOug81hI9HexkhTvgiBYGIbhwC/62hHdoxYEwelBEKwKgmBtEAQ/PMzrTgiCoCQIgvFfN6wkqXwkJgQ8cGlfhnZoxF2TlvDuysyoI6k8JSRA59Fw2Ytw51I4+W7I/Bheugwe7A3TfwP7t0edUpL0NX3liloQBInAamAMsBX4ALgsDMPlX/C6KUA+MDEMw0mHe19X1CTp+DiQX8Rlj89lbdZBnr9+MAPbNYw6kipKSTGsfiu2yrZuKgSJ0OWM2Cpbh1Ni5U6SFDeOdUVtELA2DMP1YRgWAi8B533B674FvAxkfe2kkqRyVyctmae/MYjm9Wpw3dMfsGrngagjqaIkJkG3s+GqV+Dbi2HYt2DzHHj+Qvhjf5j9IOTmRJ1SknQEjqSotQS2fObx1rLnPhEEQUvgAuDR8osmSSovjWun8ux1g6iRksjVE+exZXde1JFU0Rp2gDE/h++ugHFPQt0WMOWn8Ptu8PINsOn92H1tkqS4dCRFLfiC5z7/L/sDwN1hGJYc9o2C4MYgCBYEQbAgOzv7SDNKkspB64Y1efa6wRwqLOHqifPJOVgQdSQdD0mp0Gs8fOOfcOu82DbI1ZPhqTPg4SEw7V7YugBKD/ufcEnScXYk96gNBX4WhuFpZY/vAQjD8N7PvGYDnxa6xkAecGMYhq992ft6j5okRWPhpt1c8cQ8OjSuzZ9uHEK9GslRR9LxVpgHy16BRc/BlnlACDUaQqdR0GlM7HOtxlGnlKQq73D3qB1JUUsiNkxkFLCN2DCRy8MwXPYlr38aeMNhIpIUv2aszuaGZz6gZ8t6PH/9YGqlOsq92srbHRvzv/ad2EduNhBAi76x0tZ5DLQcAAmJUSeVpCrnmIpa2RucSWx7YyKxiY6/CoLgZoAwDB/93GufxqImSXHvrY93cNuLixncviETrz2BtGR/EK/2Skth50ew5h1YOwW2fgBhKdRoAB1P/XS1rXaTqJNKUpVwzEWtIljUJCl6ryzaynf/8hGjujbh0asGkJzo+HZ9Rt5uWD+trLi9A7llg52b942ttHUqW23zcG1J+losapKkL/Xc3E385LWPObt3cx68tB+JCV80Q0rVXmkpZC6FNVNipW3LfAhLIK0+dDwFupwJ3c6B5BpRJ5WkSuNwRc1fgUlSNXfVkLbkFRRz75srqZmSyK8v7E2CZU2fl5AAzfvEPkZ8Hw7tgfXTP11tW/ZqbItkv6tikyUbto86sSRVahY1SRI3ndyR3IJi/vDuWmqlJvHTs7sTBJY1HUaNBtDjgthHGMLGWTD/cZjzELz/R+g0GgZ9M/bZQSSSdNQsapIkAL4zJoODBSVMnL2B2qlJfG9sl6gjqbIIAmg/IvaxfzssfAYWPg0vXgz128RW2PpdDbUaRZ1UkioN71GTJH0iDEPueWUpL32whR+e0ZWbT+4YdSRVViVFsPIN+ODJ2GpbYmps9e2EG6DVwFi5k6RqznvUJElHJAgCfnVBL/IKS/j1myuplZLIVUPbRR1LlVFi8qdbI7NWwgdPwEcvwZKXoFnv2LbInuMhpWbUSSUpLrmiJkn6D0Ulpdzy/CLeWZHJ7y7qw7gBraKOpKqg4AAs+UustGUth7R60PcKGHg9NO4UdTpJOu4czy9JOmr5RSVc/8wHzFm3i4cu788ZvZpHHUlVRRjC5jmxwrb8dSgthg6nxLZFZpzuuWySqo3DFTVPNpUkfaG05EQev3og/do04NsvLWbaqqyoI6mqCAJoOwzGT4TvLIdTfgw5q+HPV8CDfWDRc7EyJ0nVmEVNkvSlaqYkMfHaE8hoWoebn1vI3PW7oo6kqqZOUzj5LrhjCVzyAtRrCX+7PTYxcv/2qNNJUmQsapKkw6pXI5lnrxtE64Y1uf7pD/hwy96oI6kqSkyCbmfDN96CM+6DDbPg4SGxASSurkmqhixqkqSv1Kh2Ks9fP5iGtVO4ZuJ8Vu7cH3UkVVUJCTD4JrhlNqR3g1dvgpeugAOZUSeTpOPKoiZJOiLN6qXx4g1DqJGcyJVPzGd99sGoI6kqa9QRvvFPGPsrWPsOPDwYlk5ydU1StWFRkyQdsdYNa/L8DYMJw5Arn5jH1j15UUdSVZaQCMNuh5vfg4Yd4OXr4a/XQG5O1MkkqcJZ1CRJR6VTk9o8e/0gDhYUc+UT88janx91JFV16Rlw3WQY9V+w6k14aDAs/1vUqSSpQlnUJElHrUeLejx93SCyDhRw5ZOWNR0HiUkw/Ltw44zYZMi/XAWTroe83VEnk6QKYVGTJH0t/ds04IlrBrJ1zyHGPzqHTbtyo46k6qBpd7hhKoz8ESx/LTYZctWbUaeSpHJnUZMkfW3DOjbmxW8O4UB+EeMemcOy7fuijqTqIDEZRt4N35wGtdLhT5fCq7fAIY+OkFR1WNQkScekb+v6/PXmoSQnBlz62FzmeSi2jpfmvWNlbcRdsOTP8PBQWPNO1KkkqVxY1CRJx6xTkzq8fMswmtRN5eqJ85my3DOvdJwkpcCpP4YbpkBqHXhhHPztW5DvWX+SKjeLmiSpXLSoX4O/3jyMrs3rcvPzC/nrgi1RR1J10nIA3DQTTrwDFj8PjwyD9dOjTiVJX5tFTZJUbhrWSuHFGwYzrGMj7pq0hMdmrIs6kqqT5DQY8wu47m1ITIFnz4M374aiQ1Enk6SjZlGTJJWrWqlJPHnNCZzduzn3vrmS//nnCsIwjDqWqpPWg2KHZA+6CeY9Co+dDNs/jDqVJB0Vi5okqdylJCXw4KX9uGpIWybMXM9dk5ZQXFIadSxVJyk14cz74MqXIX8fPDEaZv0OSkuiTiZJR8SiJkmqEIkJAb84rwd3ju7MpIVbufn5ReQX+UOyjrNOo+HWOdD1LJj6C3jqTNizMepUkvSVLGqSpAoTBAF3js7gF+f1YOrKTK6eOJ/9+UVRx1J1U7MhXPQ0XPAYZC2HR06MDRxxS66kOGZRkyRVuKuHtuPBS/uxePMeLnlsLlkH8qOOpOomCKDPpXDLbGjeF16/Df58JeTmRJ1Mkr6QRU2SdFyc26cFT15zAhtzcrno0Tls3pUXdSRVR/XbwDV/gzG/hDWTY4dkr54cdSpJ+g8WNUnScTMiI50XvzmYfYeKGPfo+6zY4aHEikBCIpz4bfjmNKiVDi9eBG98Bwpzo04mSZ+wqEmSjqt+bRow6eahJCUEXPzYHOZv2B11JFVXzXrCN9+FYd+CBU/Bo8Nh68KoU0kSYFGTJEWgU5M6TLplGOl1UrnqyXm8szwz6kiqrpLTYOx/wzV/h+ICeHIMTP81lBRHnUxSNWdRkyRFomX9Gky6eRhdm9XhpucX8tcFW6KOpOqs/fDYoJFe42H6vTBxLOxaF3UqSdWYRU2SFJmGtVJ44ZtDGNqhEXdNWsIj09cROjJdUalRHy6cAOOfipW0R0+CBRMd4y8pEhY1SVKkaqcm8eS1AzmnTwt+89ZKfvTqUopKSqOOpeqs54WxQ7JbD44NGXnxEjiYFXUqSdWMRU2SFLnUpEQevKQvt53SkT/N38J1T3/gwdiKVt0WcOUrcMZ9sGFGbNDI5nlRp5JUjVjUJElxISEh4K7TunLf+N7MWbeLcQ+/z5bdnrWmCCUkwOCbYpMhU2rC02fCvAluhZR0XFjUJElx5eKBrXn2+kFk7s/ngodns3jznqgjqbpr2iN25lqn0fDmXfDqzVDoLxEkVSyLmiQp7gzr2JhXbj2RmilJXDphLv9cuiPqSKruatSHS/8EI38ES/4MT46F3RuiTiWpCrOoSZLiUqcmtXn11mH0bFmPW19Y5ERIRS8hAUbeDVf8FfZthgkjYc2UqFNJqqIsapKkuNWodiov3DD4k4mQP3zZiZCKA53HwI3ToV5reOEimHEflHpdSipfFjVJUlxLS45NhPzWqZ3484ItXPvUfPYdciKkItawA1w/GXpfDNN+BS9dDof2Rp1KUhViUZMkxb2EhIDvje3C/Rf1Yf6G3Yx7xImQigMpNeGCx+CM38LaKfD4KZC5LOpUkqoIi5okqdIYP6AVz143mKz9+Zz/0GwWORFSUQsCGHwjXPsPKMyFJ0bD0klRp5JUBVjUJEmVytCOjXj1thOplZrEZRPm8o8lToRUHGgzBG6aCc37wMvXw1s/ghK36Er6+ixqkqRKp2P6pxMhb3txEQ9NW+tESEWvTjO45u8w+GaY+xA8ex4czIo6laRKyqImSaqU/jUR8tw+Lfjt26u4++UlFBY7eU8RS0yGM34DF0yAbYvgsRGwZX7UqSRVQhY1SVKllZacyIOX9uXbp3biLwu2xiZC5rndTHGgzyVwwxRISoWnzoQPngBXfSUdBYuaJKlSC4KA747twu8u6sMHG3dz4SOz2bzLiZCKA816xc5b63gK/ON78NqtUHQo6lSSKgmLmiSpShg3oBXPXT+YnIOFnP/wbN5flxN1JAlqNIDL/gwn/xA+ehGeHAMbZrm6JukrWdQkSVXGkA6NePXWYTSomcyVT8zjkenrHDKi6CUkwCn3xArbwSx45mx46gxY+46FTdKXsqhJkqqUDum1ef32kzijV3N+89ZKbnxuIfsOed+a4kCX0+GOj+DM+2HvFnh+XOyQ7JX/gFIH4Uj6d0FUv2kcOHBguGDBgki+tySp6gvDkKdmb+R//rmClg1q8MgVA+jeom7UsaSY4kJY8hLM+j3s2QBNesCI70H38yEhMep0ko6TIAgWhmE48Iu+5oqaJKlKCoKA605qz0s3DiG/qIQLHp7NpIVbo44lxSSlQP+r4fYFcOHjUFoMk66DhwbBhy96WLYki5okqWob2K4hb3xrOP3bNOD7f/2Ie15ZSn5RSdSxpJjEJOh9Mdw6Fy5+FpJrwGu3wB/7w4KJUFwQdUJJEXHroySpWiguKeV3U1bzyPR19GpZj4ev6E/rhjWjjiX9uzCE1W/DzN/CtgVQpwWc+G3ofw2keL1KVc3htj5a1CRJ1crkZTv53l8+IiEh4IFL+3JKlyZRR5L+UxjChhkw47ew6T2o2RiG3Q4n3ACpdaJOJ6mcWNQkSfqMjTm53Pz8QlZlHuDbp3bm26M6k5gQRB1L+mKb3oeZ98O6qZBWH4bcAoNvip3RJqlSs6hJkvQ5hwpL+H+vLeWVRdsYkZHOA5f0pWGtlKhjSV9u20KY+TtY9Q9IqQODboAht0JtV4WlysqiJknSFwjDkD/N38LP/raM9DqpPHxFf/q0rh91LOnwdn4Ms34Hy16FpNTY9Mhh34L6baJOJukoOZ5fkqQvEAQBlw9uw6RbhgJw0aNzeH7uJqL6JaZ0RJr1hIueio3273URLHgK/tAPXr0FsldFnU5SOXFFTZIkYE9uIXf++UNmrM7mwn4t+dUFvaiR4sHDqgT2bYX3/w8WPg3F+dDtHBj+XWjRL+pkkr6CWx8lSToCpaUhf3x3LQ9MXU2XpnV45MoBtG9cK+pY0pHJzYF5j8K8CVCwDzqeCsO/B21PhMBhOVI8sqhJknQUZqzO5o6XFlNSEvKb8b05s1fzqCNJRy5/Pyx4EuY8BLnZ0HpwrLB1Hmthk+KMRU2SpKO0dU8et724mI+27OXC/i352bk9qJuWHHUs6cgVHYLFz8PsP8C+zdC0J5z0HehxASS4rVeKBxY1SZK+hqKSUv44dQ0PTV9Hs7pp3H9RH4Z2bBR1LOnolBTB0knw3u8hZzU07AAn3gl9Lo1NjZQUGYuaJEnHYPHmPXz3Lx+xISeXG05qz/dP60JasisSqmRKS2HlG7HR/js+hDotYmP9B1wDKd6LKUXBoiZJ0jHKKyzm3n+u5Lm5m8hoWpvfX9yXni3rRR1LOnphCOvehff+FzbOghoNoNu50PVsaD8CktOiTihVGxY1SZLKyfRVWfxg0hL25BVy5+gMbhrRgaREjyVVJbV5Hsx7BNZMgcKDkFwLOo2KlbaMsbESJ6nCWNQkSSpHe3IL+fHrH/OPJTvo36Y+v7+4L+0c46/KrLgANsyElf+AVW/CwZ0QJELbYbHS1vVMqN8m6pRSlWNRkySpnIVhyN8+2s5PXvuY4tKQ/3dWNy4f1IbA8eeq7EpLYfuistL2T8heGXu+aa9YYet6FjTr7ah/qRxY1CRJqiA79h3irr8u4b21OZzSJZ3fjOtNk7re46MqZNe6T0vb5rlACPVaQ5czYqWt7YmQ6NEV0tdhUZMkqQKVloY8O2cj9765kpopifzPBb04w0OyVRUdzIbVb8VK27p3oTgf0urFDtPucmZsq2RqXUiu4YqbdAQsapIkHQdrsw7ynT9/yNJt+7iwX0t+dp6HZKsKK8yD9dNiq22r34K8XZ9+LUiAlNplH7Ugtfanj1PLnkupDal1vvjPaXWhSQ9ITIru7ycdBxY1SZKOk6KSUv747loemraWpnVSuf+iPgzr1DjqWFLFKi2BLfMgcxkUHIDC3NgUycKDUHDw08cFBz99vjAXivK+/D2b9IAz74N2Jx2/v4d0nFnUJEk6zj7cspfv/vlD1ufkct2J7fnB6R6SLf2H0pIvLnF7NsGM38C+LdBzHIz5JdRrGXVaqdxZ1CRJisChwhLufXMFz87ZROcmtblvfG/6tfFcKumIFObB7AfgvQcgIRFGfB+G3g5JqVEnk8rN4YqaJ3RKklRBaqQk8ovzevLsdYM4kF/MhY+8z8//vozcguKoo0nxL6UmnPIjuH0+dDwVpv4CHh4Cq9+OOpl0XFjUJEmqYCMy0pny3RFcObgtT83eyNj/ncn0VVlRx5Iqhwbt4NIX4MpXYodwv3gxvHBx7NgAqQqzqEmSdBzUSUvml+f35K83DyUtOYFrn/qAO19azO7cwqijSZVDp1Fwy/ux+9U2zY6trr3z89i9bVIV5D1qkiQdZwXFJTw0bR2PTF9LnbRkfnp2d87r24LAc6ekI3NgJ0z5L1jyEtRpAWN/GRs64v+HVMl4j5okSXEkNSmR747J4I1vDadNw5rc+ecPufapD9i65zCjyiV9qk4zuPAxuO5tqNUYXr4enj4bdn4cdTKp3LiiJklShEpKQ56bs5H73l4FwPfHduGaYe1ITHBlQDoipSWw6JnYsJH8fXDCDbEhJDWcsKr453h+SZLi3La9h/h/ry5l+qps+rSuz2/G9aJrs7pRx5Iqj7zdMO1XsGBirKSN+in0uyo22l+KU259lCQpzrWsX4Onrj2BBy/ty5bdeZz9h/f43eRV5BeVRB1NqhxqNoSzfgc3zoDGGfD3O+DxU2HT+xDRwoR0LFxRkyQpzuzOLeS//7GcVxZto0N6LX59YW8GtW8YdSyp8ghDWDoJpvwEDuyA+m2h+3nQ43xo0d+hI4obx7z1MQiC04EHgUTgiTAMf/25r58H/BIoBYqBO8MwfO9w72lRkyTp8GauzuZHry5l655DXDG4DXef0ZW6aclRx5Iqj4KDsOwVWP46rJ8OpcVQv02stHU/H1oOsLQpUsdU1IIgSARWA2OArcAHwGVhGC7/zGtqA7lhGIZBEPQG/hKGYdfDva9FTZKkr5ZXWMzvJ69m4uwNpNdJ5Zfn9WRsj2ZRx5Iqn7zdsOqfsOy1stJWBPVaf1raWg20tOm4O9aiNhT4WRiGp5U9vgcgDMN7D/P6iWEYdjvc+1rUJEk6ch9t2cvdLy9h5c4DnNGzGT89pzvN69WIOpZUOR3aA6vejJW2de/GSlvdVtD93LLSdgIkOMpBFe9Yi9p44PQwDG8oe3wVMDgMw9s/97oLgHuBJsBZYRjOOdz7WtQkSTo6RSWlTJi5nj9MXUNiQsAdozpz3UntSU70B0rpazu0N1balr8O66ZCSWHsEO3u58U+Wg+2tKnCHGtRuwg47XNFbVAYht/6ktePAH4ahuHoL/jajcCNAG3atBmwadOmo/qLSJIk2LI7j5//fRnvrMgio2ltfnleTwZ3aBR1LKnyy98Hq96C5a/B2qlQUgB1mkO3c2ODSFoNgsSkqFOqCjmuWx/LXrMBOCEMw5wve40rapIkHZspyzP52d+WsW3vIS7s15J7zuxGep3UqGNJVUP+flj9dqy0rZkSK22JqdCkGzTrBc16Q7Oe0LQHpNWLOq0qqWMtaknEhomMArYRGyZyeRiGyz7zmk7AurJhIv2BvwOtwsO8uUVNkqRjd6iwhIemreWxmetIS07k+2O7cOWQtiQmOBRBKjcFB2JlbdtCyPwYdi6FvF2ffr1+238vb816xQaVOJxEX6E8xvOfCTxAbDz/xDAMfxUEwc0AYRg+GgTB3cDVQBFwCLjL8fySJB0/67IP8l+vL+O9tTn0bFmXX57Xk35tGkQdS6qawhAO7IwVtp1LPi1vu9YBZT9bp9WDpr3KClxZeUvvCkmueutTx1zUKoJFTZKk8hWGIW8s2cF//2M5WQcKuPSENvzgtC40qJUSdTSpeijMhczl/17eMpdBUV7s6wlJ0LgLtOgLHU+NfdT0MPvqzKImSVI1crCgmAemrOap9zdSNy2JH57RlYsGtCbB7ZDS8VdaArs3/Ht527oADu2GICE2oKTzGOg8Nrbq5nbJasWiJklSNbRy535+/OrHLNi0h/5t6vPL83vSo4VDD6TIlZbA9sWwZnLsY/vi2PN1mkOn0bHS1mEkpNWNMqWOA4uaJEnVVGlpyMuLtvLrN1eyJ6+Qa4a147tjMqiTlhx1NEn/cjAL1r4TmzK5bhoU7IOEZGg7NFbaOo+FxhmutlVBFjVJkqq5fXlF/HbySl6Yt5nGtVP58VndOLdPCwJ/8JPiS0kRbJlftto2BbLKBq3Xb/NpaWs3HFJqRptT5cKiJkmSAPhoy15+8vrHLNm6j6EdGvGzc3vQpVmdqGNJ+jL7tsYK25rJsH56bDBJYiq0Hx4rbY06fWalrezz13mcmBqbTulUyuPKoiZJkj5RUhry4vzN/PatlRwsKGb8gFZ8d0wXmtVLizqapMMpLoBNsz8tbrvWlu/7J9WIbbfsMDL20bQXJCSU7/c4XkqKIWd1bIjLjiVQlAvnPBh1qv9gUZMkSf9hT24h/zdtLc/O2UhiQsA3h3fgppM7Ujs1Kepoko7E7g1wMDN2rtsnyv78yXNH+LjgAGycHVu1y14Re65GQ+hw8qfFrUG78v4blI+iQ2XHInwUK2U7PoKs5VCcH/t6Uhq06A/f+Gfc3ednUZMkSV9q86487nt7JW8s2UHj2incMTqDS09oTXJiJf1NuqRjs38HbJgZK23rp8GBHbHnG7T7tLS1PzmaM+AO7YkdcbBjyaerZTmrISyJfT2tHjTrDc37lH3uDY06Q2J8/gLKoiZJkr7Sh1v28j//WMH8jbvpkF6LH57elTHdmzpwRKrOwhBy1sQK2/rpsGEWFB4AglgJ6jAy9tFmKCTXKJ/vWVIUOzy8YP+nB4jv+Cj2ee/mT19Xp/mnZexfn+u3jbtVs8OxqEmSpCMShiHvrMji12+uYF12LoPaNeSeM7vSr02DqKNJigclxbB9Udlq2/TYhMrSotgwkjZDYqWtUUcozIvdF1aYFytd//pz0b8e5/37a4ryoPBg7M+lRf/5fRt2/PdC1qwP1E4/vn/3CmBRkyRJR6W4pJSXPtjCA++sJudgIWf1bs7dp3WlTSNHgkv6jIKDsHnOp8Ut8+Mvfl1SGiTXhJTasaMFkmtCSq2yzzUhuda/P/+vj8YZ0LRnlT3826ImSZK+loMFxUyYsY7HZ22guLSUq4a041undqJBrZSoo0mKRwezYh+fL18JiVEni0sWNUmSdEwy9+fz+8mr+evCLdRKTeL2UzpxzbB2pCX7w5ckfV2HK2qOc5IkSV+pad00fjO+N2/eMYIBbRtw75srGfW7Gby2eBulpdH80leSqjKLmiRJOmJdmtXh6W8M4oUbBlO/ZjJ3/vlDzn3oPWavzYk6miRVKRY1SZJ01E7s1Ji/334S/3tJH/bkFnHFE/O45LE5zF2/K+poklQleI+aJEk6JvlFJfxp/mYenr6O7AMFDO3QiO+MyWBQ+wgOw5WkSsRhIpIkqcLlF5XwwrzNPDJ9HTkHCzixUyPuHJ3BCe0sbJL0RSxqkiTpuDlUWMIL8zbx6Ix15Bws5KROjfnOmM4MaGthk6TPsqhJkqTj7lBhCc/PjRW2XbmFDO/cmO+MyaB/mwZRR5OkuGBRkyRJkckrLOa5OZt4bOZ6ducWcnJGOneO7kw/C5ukas6iJkmSIpdbUMxzczcxoaywjeySzp2jM+jbun7U0SQpEhY1SZIUN3ILinlmzkYmzFzP3rwiTu3ahDtHd6Z3KwubpOrFoiaVYuDqAAAWK0lEQVRJkuLOwYJinnl/I4/PihW20d2acMeoDHq1qhd1NEk6LixqkiQpbh3ILyorbBvYd6iIs3o154dndKV1w5pRR5OkCmVRkyRJcW9/fhFPztrAhJnrKQlDrj+pPbed0onaqUlRR5OkCnG4opZwvMNIkiR9kbppyXxnTAbTvj+Ss3s355Hp6xj52+n8+YPNlJRG84tlSYqKRU2SJMWVZvXS+P3FfXntthNp26gmd7+8lHP++B5z1++KOpokHTcWNUmSFJf6tq7PpJuH8sfL+rHvUBGXTpjLzc8tZPOuvKijSVKFs6hJkqS4FQQB5/RpwdTvncz3x2Ywc002o38/g3vfXMGB/KKo40lShbGoSZKkuJeWnMjtp3Zm2vdHcm7fFjw2Yz2n3D+dP833/jVJVZNFTZIkVRpN66Zx/0V9+PvtJ9G+cS3ueWUpZ/1hFu+vzYk6miSVK4uaJEmqdHq1qsdfbhrKw1f052BBMZc/MY9vPruADTm5UUeTpHJhUZMkSZVSEASc2as573z3ZH5wehfeX5vD2P+dwa/+sZx9h7x/TVLlZlGTJEmVWlpyIreO7MS0u0ZyYb9WPPHeBk65fzrPzd1EcUlp1PEk6WuxqEmSpCqhSZ00fjO+N3+//SQ6N6nNT177mNMemMnby3YShg4ckVS5WNQkSVKV0rNlPV66cQiPXz0QgJueW8j4R+ewcNPuiJNJ0pGzqEmSpConCALGdG/K23eO4N4Le7Fldx7jHpnDTc8tYF32wajjSdJXCqLaCjBw4MBwwYIFkXxvSZJUveQVFjPxvQ08OmM9h4pKuOSE1tw5ujNN6qRFHU1SNRYEwcIwDAd+4dcsapIkqbrYdbCAP767lufnbiIlKYEbhnfgxhEdqJ2aFHU0SdWQRU2SJOkzNubk8tvJq/jHkh00rp3CHaM6c+mgNiQneleIpOPncEXNf40kSVK1065xLR66vD+v3XYiHdNr85PXlzH2f2fyz6U7nBApKS5Y1CRJUrXVt3V9XrpxCBOvHUhyYsCtLyzigoffZ976XVFHk1TNWdQkSVK1FgQBp3Ztypt3jOC+cb3ZuS+fSybM5YZnPmBN5oGo40mqprxHTZIk6TMOFZYwcfYGHp2+jtzCYi4e2Jo7Rnemeb0aUUeTVMU4TESSJOko7c4t5I/vruH5uZsAuLBfK24e2ZH2jWtFnExSVWFRkyRJ+pq27M5jwsz1/HnBFopKSjmzV3NuObkjPVvWizqapErOoiZJknSMsg7kM/G9jTw/dxMHC4o5OSOd207pxKD2DaOOJqmSsqhJkiSVk32Hinh+7iYmvreBXbmFDGzbgFtP6cgpXZoQBEHU8SRVIhY1SZKkcnaosIS/LNjChJnr2bb3EF2b1eGWkR05q1dzkjw4W9IRsKhJkiRVkKKSUl7/cDuPTF/Luuxc2jaqyU0jOjJuQEtSkxKjjicpjlnUJEmSKlhpacjk5Zk8Mn0tH23dR5M6qdwwvD2XD25L7dSkqONJikMWNUmSpOMkDENmr93Fw9PX8v66XdSrkcw1w9rxjWHtaFArJep4kuKIRU2SJCkCH27Zy8PT1jJ5eSY1khO5bFAbrh/enpb1PTxbkkVNkiQpUmsyD/DIjHW8/uF2AuDs3s25cURHureoG3U0SRGyqEmSJMWBbXsPMfG9Dbw0fzO5hSUM79yYG0d04KROjR3tL1VDFjVJkqQ4si+viBfmb+Kp2RvJPlBA9+Z1uXFEB87q3ZxkR/tL1YZFTZIkKQ4VFJfw+uLtTJi1nrVZB2lZvwbfOLEdlw5q46RIqRqwqEmSJMWx0tKQaauyeGzmeuZv2E3dtCSuGNKWbwxrR5O6aVHHk1RBLGqSJEmVxOLNe3h81nre+ngnSQkJnN+vBTeO6ECnJnWijiapnFnUJEmSKplNu3J5YtYG/rJgCwXFpYzq2oQbR3RgUPuGDh6RqgiLmiRJUiW162ABz87ZxLNzNrInr4g+retz04gOnNajGYkJFjapMrOoSZIkVXKHCkuYtHALT7y3gU278mhZvwYX9m/JuP6taNe4VtTxJH0NFjVJkqQqoqQ0ZPKynfzpgy3MWpNNGMIJ7Rpw0YDWnNm7udMipUrEoiZJklQF7dh3iFcXb2PSwq2sz86lRnIiZ/RsxvgBrRjSoREJbo2U4ppFTZIkqQoLw5DFW/YyaeFW/v7Rdg7kF9Oyfg3G9W/JuAGtaNvIrZFSPLKoSZIkVRP5RSVMXp7JpIVbP9kaOah9Q8YPaMWZvdwaKcUTi5okSVI19MnWyAVbWZ9TtjWyV9nWyPZujZSiZlGTJEmqxsIwZNHm2NbINz7azoGCsq2RA1oxvn8r2jSqGXVEqVqyqEmSJAmIbY18e9lOJi3cyntrcz6ZGnl+v5ac3asF9WomRx1RqjYsapIkSfoPO/Yd4pVF23h18TbWZh0kJTGBU7qmc0G/VpzSNZ3UpMSoI0pVmkVNkiRJXyoMQ5Zt38+ri7fx+ofbyTlYQL0ayZzVuzkX9GvJwLYNCALvZ5PKm0VNkiRJR6S4pJTZ63bx6qKtvL0sk0NFJbRqUIML+rXk/H4t6ZheO+qIUpVhUZMkSdJRyy0o5u1lO3l18TZmr82hNIQ+repxfr+WnNOnBY1rp0YdUarULGqSJEk6Jln78/nbR9t5dfE2lm3fT2JCwIjOjbmgfyvGdGtKjRTvZ5OOlkVNkiRJ5WZ15oHY/WyLt7F9Xz61UhI5vWdzzuvbgiEdGpGSlBB1RKlSsKhJkiSp3JWWhszbsJtXF2/lzaU7OVBQTJ20JEZ1bcJpPZoxIiOdWqlJUceU4pZFTZIkSRUqv6iEWWtyeHvZTqauyGRPXhEpSQmM6NyYsd2bMapbExp5T5v0bw5X1PwVhyRJko5ZWnIiY7o3ZUz3phSXlPLBxj1MXr6TycsyeWdFFgkBDGzXkNN6NGNs96a0blgz6shSXHNFTZIkSRXmX2e0TV62k8nLM1m58wAA3ZvXZWyPppzWoxldm9XxnDZVS259lCRJUlzYmJPLlOWZvL1sJws37yEMoU3Dmozt3pSxPZoxoG0DEhMsbaoeLGqSJEmKO1kH8pm6Iou3l+3k/bW7KCwppVGtFEZ3a8q5ZRMkLW2qyixqkiRJimsH8ouYviqbycszmbYyi4MFxTSvl8YF/VoybkArOqbXjjqiVO4sapIkSao08otKmLI8k5cXbWXm6mxKQ+jTuj7j+7fknD4tqF8zJeqIUrk45qIWBMHpwINAIvBEGIa//tzXrwDuLnt4ELglDMOPDveeFjVJkiR9lawD+by+eDsvL9rKyp0HSE4MGNW1KeMGtGJkl3SSEz1cW5XXMRW1IAgSgdXAGGAr8AFwWRiGyz/zmmHAijAM9wRBcAbwszAMBx/ufS1qkiRJOhrLtu/jlUXbeP3DbeQcLKRhrRTO7dOC8QNa0aNFXSdHqtI51qI2lFjxOq3s8T0AYRje+yWvbwB8HIZhy8O9r0VNkiRJX0dRSSkzV2fzyqJtTFmeSWFJKRlNazOufyvO79eSpnXToo4oHZFjPfC6JbDlM4+3AodbLbseePPI40mSJElHLjkxgVHdmjKqW1P25RXxxtLtvLxwK/e+uZLfvLWSkzqnM65/S8Z2b0aNlMSo40pfy5EUtS9aQ/7CZbggCE4hVtRO+pKv3wjcCNCmTZsjjChJkiR9sXo1k7licFuuGNyWDTm5vLJoK68s2sYdL31IndQkTuvZjJFd0jmpU2OHkKhSKbetj0EQ9AZeBc4Iw3D1V31jtz5KkiSpIpSWhszbsJtXFm3l7WU72Z9fTEIQmxw5onM6IzLS6du6vme0KXLHeo9aErFhIqOAbcSGiVwehuGyz7ymDfAucHUYhu8fSSiLmiRJkipacUkpH23dx8zV2cxck81HW/ZSGkLdtCSGd05nREZjRmSk07xejaijqhoqj/H8ZwIPEBvPPzEMw18FQXAzQBiGjwZB8AQwDthU9j8p/rJv+C8WNUmSJB1ve/MKeW9tTqy4rc5h5/58ADo3qc2IjHROzkhnUPuGpCV7b5sqngdeS5IkSZ8ThiGrMw9+sto2b8NuCotLSU1KYHCHRozo3JiTM9Lp1KS2o/9VISxqkiRJ0lc4VFjC3A27ylbbslmXnQtAi3ppjMiI3ds2rGMjh5Ko3FjUJEmSpKO0dU8es9bkMGNVNrPX5XCgbChJ71b1GdG5McPLhpIkJyZEHVWVlEVNkiRJOgafHUoya002H5YNJamdmsTQjrFtksM7p9O2UU23SeqIWdQkSZKkcrTvUBFz1uUwc01sMMnWPYcAaN2wRmyaZOfGDO3YmHo1kiNOqnhmUZMkSZIqSBiGbNqVx6w12cxak8P763ZxsODTs9uGd07n5IzG9GlVnyS3SeozLGqSJEnScVJUUspHW/Yyc00Osz5zdludsm2SwzPSOblzOm0a1Yw6qiJmUZMkSZIisi+viPc/s01y297YNsm2jWoyonNsmuTQjo2onZoUcVIdbxY1SZIkKQ6EYciGnNyyoSQ5zFm/i7zCEpISAvq3bcDJGekM79yYni3qkZDgUJKqzqImSZIkxaGC4hIWbtrDrLLVtmXb9wPQoGYyJ5UNJRmRkU7TumkRJ1VFsKhJkiRJlUD2gQJmr42Vtplrcsg5WABAl6Z1GJERK20ntGtIWnJixElVHixqkiRJUiUThiErdhxg5prY2W0fbNhDYUkpqUkJDO4QO7vt5Ix0OjWp7dltlZRFTZIkSark8gqLmbd+NzPXZDNzdTbrsnMBaFm/Bid3SefkjHRO7NTYoSSViEVNkiRJqmK27T3EjFXZTF+Vxey1OeQWlpCcGDCwbUNGdklnZJcmZDR1tS2eWdQkSZKkKqywuJQFm3aXFbdsVmUeAKBFvbSy1bYmnNipEXXSkiNOqs+yqEmSJEnVyPa9h5ix+l+rbbs4WFBMUkLAwHYNGNmlCSO7pNOlaR1X2yJmUZMkSZKqqcLiUhZu2sP01VnMWJXNyp2x1bZmddM4OSOdkV3SObFzY+q62nbcWdQkSZIkAbBj3yFmro5tkXxvTQ4HylbbBrVvyKhuTRndrQltG9WKOma1YFGTJEmS9B+KSkpZtGkP01ZlM3VFJmuyDgLQuUntT0pbvzYNSExwi2RFsKhJkiRJ+kqbduXyzoospq7IZP6G3RSXhjSslcIpXZowpnsThndOp5bj/8uNRU2SJEnSUdl3qIgZq2MrbdNWZrE/v5iUxASGdGzEmG5NGNWtKS3q14g6ZqVmUZMkSZL0tRWVlLJg4x6mrsjknRWZbNyVB0C35nU/KW29WtYjwS2SR8WiJkmSJKlchGHIuuxcpq7IZOqKLBZs2k1pCOl1UhnVtQmndm3C0I6e2XYkLGqSJEmSKsTu3EKmr8pi6oosZqzO/uTMtv5tGjC8c2OGZ6TTq2U9B5J8AYuaJEmSpAr3rzPbZq3JZtaaHJZu2wdAvRrJnNSp8SfFraX3tgEWNUmSJEkR2HWwgNnrdjFzdTaz1mSTub8AgI7ptRjeOZ0RGY0Z3L5RtZ0kaVGTJEmSFKkwDFmTdbCstOUwb8Mu8otKSU4MGNC2ASMy0hnROZ3uzetWm6EkFjVJkiRJcSW/qIQFG2PbJGeuyWHFjv0ANKyV8sk2yZFdmpBeJzXipBXHoiZJkiQprmUdyGf22hxmrc5h5poccg4WEATQr3V9RndvyphuTenUpDZBUHVW2yxqkiRJkiqN0tKQ5Tv2M3VFFu+syPxkKEnbRjUZ3a0po7s15YR2DUhKTIg46bGxqEmSJEmqtHbsO/RJaXt/7S4KS0qpVyOZU7s2YXS3pozIaFwpz22zqEmSJEmqEnILipm1Jpspy7N4d2Ume/KKSE4MGNKhEWO6N2VUt6aVZvy/RU2SJElSlVNSGrJo8x7eWZ7JlBWZrM/OBaB787qf3NfWs2XduL2vzaImSZIkqcpbl32QqSsymbI8k4Wb9lAaQrO6aYzu3oSfndMj7u5pO1xRq54ny0mSJEmqcjqm16Zjem1uHNGRXQcLmLYqm3eWZ7Jyx4G4K2lfxaImSZIkqcppVDuV8QNaMX5AK6LaRXgsKletlCRJkqSjFK/3qB2ORU2SJEmS4oxFTZIkSZLijEVNkiRJkuKMRU2SJEmS4oxFTZIkSZLijEVNkiRJkuKMRU2SJEmS4oxFTZIkSZLijEVNkiRJkuKMRU2SJEmS4oxFTZIkSZLijEVNkiRJkuKMRU2SJEmS4oxFTZIkSZLijEVNkiRJkuKMRU2SJEmS4oxFTZIkSZLijEVNkiRJkuKMRU2SJEmS4oxFTZIkSZLijEVNkiRJkuKMRU2SJEmS4kwQhmE03zgIsoFN5fiWjYGccnw/Kd54jas68DpXVec1rurA6/zItQ3DMP2LvhBZUStvQRAsCMNwYNQ5pIriNa7qwOtcVZ3XuKoDr/Py4dZHSZIkSYozFjVJkiRJijNVqahNiDqAVMG8xlUdeJ2rqvMaV3XgdV4Oqsw9apIkSZJUVVSlFTVJkiRJqhIqfVELguD0IAhWBUGwNgiCH0adRyoPQRBMDIIgKwiCjz/zXMMgCKYEQbCm7HODKDNKxyIIgtZBEEwLgmBFEATLgiC4o+x5r3NVGUEQpAVBMD8Igo/KrvOflz3vda4qJQiCxCAIFgdB8EbZY6/xclCpi1oQBInAQ8AZQHfgsiAIukebSioXTwOnf+65HwJTwzDsDEwteyxVVsXA98Iw7AYMAW4r+/fb61xVSQFwahiGfYC+wOlBEAzB61xVzx3Ais889hovB5W6qAGDgLVhGK4Pw7AQeAk4L+JM0jELw3AmsPtzT58HPFP252eA849rKKkchWG4IwzDRWV/PkDsP/At8TpXFRLGHCx7mFz2EeJ1riokCIJWwFnAE5952mu8HFT2otYS2PKZx1vLnpOqoqZhGO6A2A+5QJOI80jlIgiCdkA/YB5e56piyraEfQhkAVPCMPQ6V1XzAPADoPQzz3mNl4PKXtSCL3jOMZaSVEkEQVAbeBm4MwzD/VHnkcpbGIYlYRj2BVoBg4Ig6Bl1Jqm8BEFwNpAVhuHCqLNURZW9qG0FWn/mcStge0RZpIqWGQRBc4Cyz1kR55GOSRAEycRK2gthGL5S9rTXuaqkMAz3AtOJ3X/sda6q4kTg3CAINhK7BenUIAiex2u8XFT2ovYB0DkIgvZBEKQAlwJ/iziTVFH+BlxT9uf/397dolQURVEAXpuXbILaRByBIzC8YBKrTXASFi2CYHUGWoWXdAIWh2DQbHMU23AFxarw7rt8Xzpwywkr3HU4P6dJHpc4F/iTqqokt0leu/vmxyc5ZzKqaquq1r/Ga0kOkrxFzpmI7j7v7u3u3s3wH/7U3SeR8X+x8g9eV9Vhhr2xsyR33X295CnBn1XVfZJ5ks0kH0kukzwkWSTZSfKe5Li7f184AiuhqvaTPCd5yfe5hosM59TknEmoqr0MFynMMiyOL7r7qqo2IudMTFXNk5x195GM/4+VL2oAAABTs+pbHwEAACZHUQMAABgZRQ0AAGBkFDUAAICRUdQAAABGRlEDAAAYGUUNAABgZBQ1AACAkfkEcGbbFIhHrgAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "# Plot the train/validation loss values\n",
    "plt.figure(figsize=(15,10))\n",
    "_loss = history.history['loss'][1:]\n",
    "_val_loss = history.history['val_loss'][1:]\n",
    "\n",
    "train_loss_plot, = plt.plot(range(1, len(_loss)+1), _loss, label='Train Loss')\n",
    "val_loss_plot, = plt.plot(range(1, len(_val_loss)+1), _val_loss, label='Validation Loss')\n",
    "\n",
    "_ = plt.legend(handles=[train_loss_plot, val_loss_plot])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Test the model\n",
    "\n",
    "After training on the train data and validating on the validation data, we test the final prediction power of our model by running it on the test dataset that the algorithm has NEVER seen before.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "78/78 [==============================] - 0s 109us/sample - loss: 0.3087 - accuracy: 0.9103\n"
     ]
    }
   ],
   "source": [
    "test_loss, test_accuracy = model.evaluate(test_inputs, test_targets)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Apply Deep Learning Model to hthe unbalanced dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [],
   "source": [
    "#first lets convert the inputs to array\n",
    "y_train1 = np.array(y_train1)\n",
    "y_test1 = np.array(y_test1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 148773 samples, validate on 49592 samples\n",
      "Epoch 1/100\n",
      "148773/148773 - 5s - loss: 0.0967 - accuracy: 0.9980 - val_loss: 0.0100 - val_accuracy: 0.9983\n",
      "Epoch 2/100\n",
      "148773/148773 - 3s - loss: 0.0117 - accuracy: 0.9980 - val_loss: 0.0092 - val_accuracy: 0.9983\n",
      "Epoch 3/100\n",
      "148773/148773 - 3s - loss: 0.0106 - accuracy: 0.9980 - val_loss: 0.0081 - val_accuracy: 0.9983\n",
      "Epoch 4/100\n",
      "148773/148773 - 4s - loss: 0.0093 - accuracy: 0.9980 - val_loss: 0.0078 - val_accuracy: 0.9983\n",
      "Epoch 5/100\n",
      "148773/148773 - 4s - loss: 0.0080 - accuracy: 0.9980 - val_loss: 0.0051 - val_accuracy: 0.9983\n",
      "Epoch 6/100\n",
      "148773/148773 - 5s - loss: 0.0046 - accuracy: 0.9981 - val_loss: 0.0035 - val_accuracy: 0.9993\n",
      "Epoch 7/100\n",
      "148773/148773 - 4s - loss: 0.0036 - accuracy: 0.9993 - val_loss: 0.0034 - val_accuracy: 0.9993\n",
      "Epoch 8/100\n",
      "148773/148773 - 4s - loss: 0.0035 - accuracy: 0.9993 - val_loss: 0.0034 - val_accuracy: 0.9994\n",
      "Epoch 9/100\n",
      "148773/148773 - 4s - loss: 0.0034 - accuracy: 0.9993 - val_loss: 0.0033 - val_accuracy: 0.9994\n",
      "Epoch 10/100\n",
      "148773/148773 - 4s - loss: 0.0034 - accuracy: 0.9993 - val_loss: 0.0033 - val_accuracy: 0.9994\n",
      "Epoch 11/100\n",
      "148773/148773 - 4s - loss: 0.0033 - accuracy: 0.9993 - val_loss: 0.0034 - val_accuracy: 0.9993\n"
     ]
    }
   ],
   "source": [
    "# Set the input and output sizes\n",
    "input_size = 53\n",
    "output_size = 2\n",
    "# Use same hidden layer size for both hidden layers. Not a necessity.\n",
    "hidden_layer_size = 2\n",
    "    \n",
    "# define how the model will look like\n",
    "tfk = tf.keras.Sequential([\n",
    "    # tf.keras.layers.Dense is basically implementing: output = activation(dot(input, weight) + bias)\n",
    "    # it takes several arguments, but the most important ones for us are the hidden_layer_size and the activation function\n",
    "    tf.keras.layers.Dense(hidden_layer_size, activation='relu'), # 1st hidden layer\n",
    "    tf.keras.layers.Dense(hidden_layer_size, activation='relu'), # 2nd hidden layer\n",
    "    tf.keras.layers.Dense(hidden_layer_size, activation='relu'), # 2nd hidden layer\n",
    "    tf.keras.layers.Dense(hidden_layer_size, activation='relu'), # 2nd hidden layer\n",
    "    tf.keras.layers.Dense(hidden_layer_size, activation='relu'), # 2nd hidden layer\n",
    "    # the final layer is no different, we just make sure to activate it with softmax\n",
    "    tf.keras.layers.Dense(output_size, activation='softmax') # output layer\n",
    "])\n",
    "\n",
    "\n",
    "### Choose the optimizer and the loss function\n",
    "\n",
    "# we define the optimizer we'd like to use, \n",
    "# the loss function, \n",
    "# and the metrics we are interested in obtaining at each iteration\n",
    "tfk.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy'])\n",
    "\n",
    "### Training\n",
    "# That's where we train the model we have built.\n",
    "\n",
    "# set the batch size\n",
    "batch_size = 100\n",
    "\n",
    "# set a maximum number of training epochs\n",
    "max_epochs = 100\n",
    "\n",
    "# set an early stopping mechanism\n",
    "# let's set patience=2, to be a bit tolerant against random validation loss increases\n",
    "early_stopping = tf.keras.callbacks.EarlyStopping(patience=2)\n",
    "\n",
    "# fit the model\n",
    "# note that this time the train, validation and test data are not iterable\n",
    "history = tfk.fit(x_train1, # train inputs\n",
    "          y_train1, # train targets\n",
    "          batch_size=batch_size, # batch size\n",
    "          epochs=max_epochs, # epochs that we will train for (assuming early stopping doesn't kick in)\n",
    "          # callbacks are functions called by a task when a task is completed\n",
    "          # task here is to check if val_loss is increasing\n",
    "          callbacks=[early_stopping], # early stopping\n",
    "          validation_data=(x_test1, y_test1), # validation data\n",
    "          verbose = 2 # making sure we get enough information about the training process\n",
    "          )  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "# Plot the train/validation loss values\n",
    "plt.figure(figsize=(20,10))\n",
    "_loss = history.history['loss'][1:]\n",
    "_val_loss = history.history['val_loss'][1:]\n",
    "\n",
    "train_loss_plot, = plt.plot(range(1, len(_loss)+1), _loss, label='Train Loss')\n",
    "val_loss_plot, = plt.plot(range(1, len(_val_loss)+1), _val_loss, label='Validation Loss')\n",
    "\n",
    "_ = plt.legend(handles=[train_loss_plot, val_loss_plot])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Make Prediction on Test data\n",
    "Apply afew of the models to the test data to make preictions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {},
   "outputs": [],
   "source": [
    "tfbalanced = model.predict_classes(test_inputsx) # Deep learning with balanced data\n",
    "tfunbalanced = tfk.predict_classes(test_inputsx)  # Deep Learning with unbalanced data\n",
    "logreg = mod4.predict(test_inputsx)              # Logistic regression with balanced data\n",
    "ada = mod6.predict(test_inputsx)            # Adaboost with balanced data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {},
   "outputs": [],
   "source": [
    "#df = pd.DataFrame(pred, columns=['Prediction'])\n",
    "submission['tfb'] = tfbalanced          # the AUC score = 54\n",
    "submission['tfu'] = tfunbalanced        # the AUC score = 0.89\n",
    "submission['logr'] = logreg             # the AUC score = 57\n",
    "submission['ada'] = ada                 # the AUC score = 63"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {},
   "outputs": [],
   "source": [
    "submission.to_csv('submission.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The best model seem to be the tfk model,\n",
    "which is the model with the unbalanced dataset\n",
    "The model generated an AUC score of 0.89277.\n",
    "So this project gives different ways of adressing the task, with balanced and unbalanced data.\n",
    "Using deep learning, KNN, SVC, Adaboost, SGboostClassifier, Logistic Regression, GaussianNB and Random Foret"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}