Mudit_model_development.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 5,
  "metadata": {
    "colab": {
      "name": "Technocolab Team AMudit.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.8.8"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "879d5988"
      },
      "source": [
        "# Exploratory Data Analysis"
      ],
      "id": "879d5988"
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mr_2kR018dus",
        "outputId": "4b888901-341d-435a-d59e-e82f1113167b"
      },
      "source": [
        "! git clone https://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@github.com/COOLMudi/CORPORATE-BANKRUPTCY-PREDICTION.git"
      ],
      "id": "mr_2kR018dus",
      "execution_count": 75,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "fatal: destination path 'CORPORATE-BANKRUPTCY-PREDICTION' already exists and is not an empty directory.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "d35a282a"
      },
      "source": [
        "## Importing Libraries"
      ],
      "id": "d35a282a"
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0cce91ef"
      },
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "from scipy.io import arff\n",
        "import missingno as msno\n",
        "%matplotlib inline\n",
        "import warnings\n",
        "warnings.filterwarnings(\"ignore\")\n",
        "\n",
        "\n",
        "from imblearn.over_sampling import SMOTE\n",
        "\n",
        "from sklearn.feature_selection import RFE,RFECV\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.model_selection import cross_val_score\n",
        "from sklearn.pipeline import Pipeline\n",
        "from sklearn.metrics import confusion_matrix, roc_auc_score ,roc_curve,auc\n",
        "from sklearn.model_selection import GridSearchCV,StratifiedKFold"
      ],
      "id": "0cce91ef",
      "execution_count": 76,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5c18ca72"
      },
      "source": [
        "## Importing and Organizing the data"
      ],
      "id": "5c18ca72"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "e56f8206"
      },
      "source": [
        "Note:\n",
        "\n",
        "In model development all the steps carried out in EDA for data cleaning will be same, however we will take 1,2,3,5 year data for training and 4th year data for testing."
      ],
      "id": "e56f8206"
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "f502be7c"
      },
      "source": [
        "dataset_1 = arff.loadarff('/content/CORPORATE-BANKRUPTCY-PREDICTION/data/1year.arff')\n",
        "dataset_2 = arff.loadarff('/content/CORPORATE-BANKRUPTCY-PREDICTION/data/2year.arff')\n",
        "dataset_3 = arff.loadarff('/content/CORPORATE-BANKRUPTCY-PREDICTION/data/3year.arff')\n",
        "dataset_4 = arff.loadarff('/content/CORPORATE-BANKRUPTCY-PREDICTION/data/4year.arff')\n",
        "dataset_5 = arff.loadarff('/content/CORPORATE-BANKRUPTCY-PREDICTION/data/5year.arff')"
      ],
      "id": "f502be7c",
      "execution_count": 77,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "47eb7d2a"
      },
      "source": [
        "#Training dataset\n",
        "df1 = pd.DataFrame(dataset_1[0])\n",
        "df2 = pd.DataFrame(dataset_2[0])\n",
        "df3 = pd.DataFrame(dataset_3[0])\n",
        "df5 = pd.DataFrame(dataset_5[0])\n",
        "\n",
        "#Testing dataset\n",
        "df4 = pd.DataFrame(dataset_4[0])"
      ],
      "id": "47eb7d2a",
      "execution_count": 78,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "19507e67"
      },
      "source": [
        "df_train = pd.concat([df1,df2,df3,df5],axis = 0)\n",
        "\n",
        "df_test = df4"
      ],
      "id": "19507e67",
      "execution_count": 79,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bfa6c81f",
        "outputId": "cb1d4e3c-1b0b-42f8-db13-9d66748dab10"
      },
      "source": [
        "print(\"{0} Rows and {1} columns for Training dataset\".format(df_train.shape[0],df_train.shape[1]))\n",
        "print(\"{0} Rows and {1} columns for Test dataset\".format(df_test.shape[0],df_test.shape[1]))"
      ],
      "id": "bfa6c81f",
      "execution_count": 80,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "33613 Rows and 65 columns for Training dataset\n",
            "9792 Rows and 65 columns for Test dataset\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 224
        },
        "id": "dc7e8e3b",
        "outputId": "aaa0cba6-c976-403e-f249-80c79bf059d9"
      },
      "source": [
        "#Top five records in the Train data\n",
        "df_train.head()"
      ],
      "id": "dc7e8e3b",
      "execution_count": 81,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>Attr1</th>\n",
              "      <th>Attr2</th>\n",
              "      <th>Attr3</th>\n",
              "      <th>Attr4</th>\n",
              "      <th>Attr5</th>\n",
              "      <th>Attr6</th>\n",
              "      <th>Attr7</th>\n",
              "      <th>Attr8</th>\n",
              "      <th>Attr9</th>\n",
              "      <th>Attr10</th>\n",
              "      <th>Attr11</th>\n",
              "      <th>Attr12</th>\n",
              "      <th>Attr13</th>\n",
              "      <th>Attr14</th>\n",
              "      <th>Attr15</th>\n",
              "      <th>Attr16</th>\n",
              "      <th>Attr17</th>\n",
              "      <th>Attr18</th>\n",
              "      <th>Attr19</th>\n",
              "      <th>Attr20</th>\n",
              "      <th>Attr21</th>\n",
              "      <th>Attr22</th>\n",
              "      <th>Attr23</th>\n",
              "      <th>Attr24</th>\n",
              "      <th>Attr25</th>\n",
              "      <th>Attr26</th>\n",
              "      <th>Attr27</th>\n",
              "      <th>Attr28</th>\n",
              "      <th>Attr29</th>\n",
              "      <th>Attr30</th>\n",
              "      <th>Attr31</th>\n",
              "      <th>Attr32</th>\n",
              "      <th>Attr33</th>\n",
              "      <th>Attr34</th>\n",
              "      <th>Attr35</th>\n",
              "      <th>Attr36</th>\n",
              "      <th>Attr37</th>\n",
              "      <th>Attr38</th>\n",
              "      <th>Attr39</th>\n",
              "      <th>Attr40</th>\n",
              "      <th>Attr41</th>\n",
              "      <th>Attr42</th>\n",
              "      <th>Attr43</th>\n",
              "      <th>Attr44</th>\n",
              "      <th>Attr45</th>\n",
              "      <th>Attr46</th>\n",
              "      <th>Attr47</th>\n",
              "      <th>Attr48</th>\n",
              "      <th>Attr49</th>\n",
              "      <th>Attr50</th>\n",
              "      <th>Attr51</th>\n",
              "      <th>Attr52</th>\n",
              "      <th>Attr53</th>\n",
              "      <th>Attr54</th>\n",
              "      <th>Attr55</th>\n",
              "      <th>Attr56</th>\n",
              "      <th>Attr57</th>\n",
              "      <th>Attr58</th>\n",
              "      <th>Attr59</th>\n",
              "      <th>Attr60</th>\n",
              "      <th>Attr61</th>\n",
              "      <th>Attr62</th>\n",
              "      <th>Attr63</th>\n",
              "      <th>Attr64</th>\n",
              "      <th>class</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.200550</td>\n",
              "      <td>0.37951</td>\n",
              "      <td>0.39641</td>\n",
              "      <td>2.0472</td>\n",
              "      <td>32.3510</td>\n",
              "      <td>0.38825</td>\n",
              "      <td>0.249760</td>\n",
              "      <td>1.33050</td>\n",
              "      <td>1.1389</td>\n",
              "      <td>0.50494</td>\n",
              "      <td>0.249760</td>\n",
              "      <td>0.65980</td>\n",
              "      <td>0.166600</td>\n",
              "      <td>0.249760</td>\n",
              "      <td>497.42</td>\n",
              "      <td>0.73378</td>\n",
              "      <td>2.6349</td>\n",
              "      <td>0.249760</td>\n",
              "      <td>0.149420</td>\n",
              "      <td>43.370</td>\n",
              "      <td>1.2479</td>\n",
              "      <td>0.21402</td>\n",
              "      <td>0.119980</td>\n",
              "      <td>0.47706</td>\n",
              "      <td>0.50494</td>\n",
              "      <td>0.60411</td>\n",
              "      <td>1.45820</td>\n",
              "      <td>1.7615</td>\n",
              "      <td>5.9443</td>\n",
              "      <td>0.11788</td>\n",
              "      <td>0.149420</td>\n",
              "      <td>94.14</td>\n",
              "      <td>3.8772</td>\n",
              "      <td>0.56393</td>\n",
              "      <td>0.21402</td>\n",
              "      <td>1.7410</td>\n",
              "      <td>593.2700</td>\n",
              "      <td>0.50591</td>\n",
              "      <td>0.128040</td>\n",
              "      <td>0.662950</td>\n",
              "      <td>0.051402</td>\n",
              "      <td>0.128040</td>\n",
              "      <td>114.42</td>\n",
              "      <td>71.050</td>\n",
              "      <td>1.00970</td>\n",
              "      <td>1.52250</td>\n",
              "      <td>49.394</td>\n",
              "      <td>0.185300</td>\n",
              "      <td>0.110850</td>\n",
              "      <td>2.0420</td>\n",
              "      <td>0.37854</td>\n",
              "      <td>0.25792</td>\n",
              "      <td>2.2437</td>\n",
              "      <td>2.2480</td>\n",
              "      <td>348690.0</td>\n",
              "      <td>0.121960</td>\n",
              "      <td>0.39718</td>\n",
              "      <td>0.87804</td>\n",
              "      <td>0.001924</td>\n",
              "      <td>8.4160</td>\n",
              "      <td>5.1372</td>\n",
              "      <td>82.658</td>\n",
              "      <td>4.4158</td>\n",
              "      <td>7.4277</td>\n",
              "      <td>b'0'</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.209120</td>\n",
              "      <td>0.49988</td>\n",
              "      <td>0.47225</td>\n",
              "      <td>1.9447</td>\n",
              "      <td>14.7860</td>\n",
              "      <td>0.00000</td>\n",
              "      <td>0.258340</td>\n",
              "      <td>0.99601</td>\n",
              "      <td>1.6996</td>\n",
              "      <td>0.49788</td>\n",
              "      <td>0.261140</td>\n",
              "      <td>0.51680</td>\n",
              "      <td>0.158350</td>\n",
              "      <td>0.258340</td>\n",
              "      <td>677.96</td>\n",
              "      <td>0.53838</td>\n",
              "      <td>2.0005</td>\n",
              "      <td>0.258340</td>\n",
              "      <td>0.152000</td>\n",
              "      <td>87.981</td>\n",
              "      <td>1.4293</td>\n",
              "      <td>0.24806</td>\n",
              "      <td>0.123040</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.39542</td>\n",
              "      <td>0.43992</td>\n",
              "      <td>88.44400</td>\n",
              "      <td>16.9460</td>\n",
              "      <td>3.6884</td>\n",
              "      <td>0.26969</td>\n",
              "      <td>0.152000</td>\n",
              "      <td>122.17</td>\n",
              "      <td>2.9876</td>\n",
              "      <td>2.98760</td>\n",
              "      <td>0.20616</td>\n",
              "      <td>1.6996</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.49788</td>\n",
              "      <td>0.121300</td>\n",
              "      <td>0.086422</td>\n",
              "      <td>0.064371</td>\n",
              "      <td>0.145950</td>\n",
              "      <td>199.49</td>\n",
              "      <td>111.510</td>\n",
              "      <td>0.51045</td>\n",
              "      <td>1.12520</td>\n",
              "      <td>100.130</td>\n",
              "      <td>0.237270</td>\n",
              "      <td>0.139610</td>\n",
              "      <td>1.9447</td>\n",
              "      <td>0.49988</td>\n",
              "      <td>0.33472</td>\n",
              "      <td>17.8660</td>\n",
              "      <td>17.8660</td>\n",
              "      <td>2304.6</td>\n",
              "      <td>0.121300</td>\n",
              "      <td>0.42002</td>\n",
              "      <td>0.85300</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>4.1486</td>\n",
              "      <td>3.2732</td>\n",
              "      <td>107.350</td>\n",
              "      <td>3.4000</td>\n",
              "      <td>60.9870</td>\n",
              "      <td>b'0'</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.248660</td>\n",
              "      <td>0.69592</td>\n",
              "      <td>0.26713</td>\n",
              "      <td>1.5548</td>\n",
              "      <td>-1.1523</td>\n",
              "      <td>0.00000</td>\n",
              "      <td>0.309060</td>\n",
              "      <td>0.43695</td>\n",
              "      <td>1.3090</td>\n",
              "      <td>0.30408</td>\n",
              "      <td>0.312580</td>\n",
              "      <td>0.64184</td>\n",
              "      <td>0.244350</td>\n",
              "      <td>0.309060</td>\n",
              "      <td>794.16</td>\n",
              "      <td>0.45961</td>\n",
              "      <td>1.4369</td>\n",
              "      <td>0.309060</td>\n",
              "      <td>0.236100</td>\n",
              "      <td>73.133</td>\n",
              "      <td>1.4283</td>\n",
              "      <td>0.30260</td>\n",
              "      <td>0.189960</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.28932</td>\n",
              "      <td>0.37282</td>\n",
              "      <td>86.01100</td>\n",
              "      <td>1.0627</td>\n",
              "      <td>4.3749</td>\n",
              "      <td>0.41929</td>\n",
              "      <td>0.238150</td>\n",
              "      <td>176.93</td>\n",
              "      <td>2.0630</td>\n",
              "      <td>1.42740</td>\n",
              "      <td>0.31565</td>\n",
              "      <td>1.3090</td>\n",
              "      <td>2.3019</td>\n",
              "      <td>0.51537</td>\n",
              "      <td>0.241140</td>\n",
              "      <td>0.322020</td>\n",
              "      <td>0.074020</td>\n",
              "      <td>0.231170</td>\n",
              "      <td>165.51</td>\n",
              "      <td>92.381</td>\n",
              "      <td>0.94807</td>\n",
              "      <td>1.01010</td>\n",
              "      <td>96.372</td>\n",
              "      <td>0.291810</td>\n",
              "      <td>0.222930</td>\n",
              "      <td>1.0758</td>\n",
              "      <td>0.48152</td>\n",
              "      <td>0.48474</td>\n",
              "      <td>1.2098</td>\n",
              "      <td>2.0504</td>\n",
              "      <td>6332.7</td>\n",
              "      <td>0.241140</td>\n",
              "      <td>0.81774</td>\n",
              "      <td>0.76599</td>\n",
              "      <td>0.694840</td>\n",
              "      <td>4.9909</td>\n",
              "      <td>3.9510</td>\n",
              "      <td>134.270</td>\n",
              "      <td>2.7185</td>\n",
              "      <td>5.2078</td>\n",
              "      <td>b'0'</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.081483</td>\n",
              "      <td>0.30734</td>\n",
              "      <td>0.45879</td>\n",
              "      <td>2.4928</td>\n",
              "      <td>51.9520</td>\n",
              "      <td>0.14988</td>\n",
              "      <td>0.092704</td>\n",
              "      <td>1.86610</td>\n",
              "      <td>1.0571</td>\n",
              "      <td>0.57353</td>\n",
              "      <td>0.092704</td>\n",
              "      <td>0.30163</td>\n",
              "      <td>0.094257</td>\n",
              "      <td>0.092704</td>\n",
              "      <td>917.01</td>\n",
              "      <td>0.39803</td>\n",
              "      <td>3.2537</td>\n",
              "      <td>0.092704</td>\n",
              "      <td>0.071428</td>\n",
              "      <td>79.788</td>\n",
              "      <td>1.5069</td>\n",
              "      <td>0.11550</td>\n",
              "      <td>0.062782</td>\n",
              "      <td>0.17193</td>\n",
              "      <td>0.57353</td>\n",
              "      <td>0.36152</td>\n",
              "      <td>0.94076</td>\n",
              "      <td>1.9618</td>\n",
              "      <td>4.6511</td>\n",
              "      <td>0.14343</td>\n",
              "      <td>0.071428</td>\n",
              "      <td>91.37</td>\n",
              "      <td>3.9948</td>\n",
              "      <td>0.37581</td>\n",
              "      <td>0.11550</td>\n",
              "      <td>1.3562</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.57353</td>\n",
              "      <td>0.088995</td>\n",
              "      <td>0.401390</td>\n",
              "      <td>0.069622</td>\n",
              "      <td>0.088995</td>\n",
              "      <td>180.77</td>\n",
              "      <td>100.980</td>\n",
              "      <td>0.28720</td>\n",
              "      <td>1.56960</td>\n",
              "      <td>84.344</td>\n",
              "      <td>0.085874</td>\n",
              "      <td>0.066165</td>\n",
              "      <td>2.4928</td>\n",
              "      <td>0.30734</td>\n",
              "      <td>0.25033</td>\n",
              "      <td>2.4524</td>\n",
              "      <td>2.4524</td>\n",
              "      <td>20545.0</td>\n",
              "      <td>0.054015</td>\n",
              "      <td>0.14207</td>\n",
              "      <td>0.94598</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>4.5746</td>\n",
              "      <td>3.6147</td>\n",
              "      <td>86.435</td>\n",
              "      <td>4.2228</td>\n",
              "      <td>5.5497</td>\n",
              "      <td>b'0'</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.187320</td>\n",
              "      <td>0.61323</td>\n",
              "      <td>0.22960</td>\n",
              "      <td>1.4063</td>\n",
              "      <td>-7.3128</td>\n",
              "      <td>0.18732</td>\n",
              "      <td>0.187320</td>\n",
              "      <td>0.63070</td>\n",
              "      <td>1.1559</td>\n",
              "      <td>0.38677</td>\n",
              "      <td>0.187320</td>\n",
              "      <td>0.33147</td>\n",
              "      <td>0.121820</td>\n",
              "      <td>0.187320</td>\n",
              "      <td>1133.20</td>\n",
              "      <td>0.32211</td>\n",
              "      <td>1.6307</td>\n",
              "      <td>0.187320</td>\n",
              "      <td>0.115530</td>\n",
              "      <td>57.045</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.19832</td>\n",
              "      <td>0.115530</td>\n",
              "      <td>0.18732</td>\n",
              "      <td>0.38677</td>\n",
              "      <td>0.32211</td>\n",
              "      <td>1.41380</td>\n",
              "      <td>1.1184</td>\n",
              "      <td>4.1424</td>\n",
              "      <td>0.27884</td>\n",
              "      <td>0.115530</td>\n",
              "      <td>147.04</td>\n",
              "      <td>2.4823</td>\n",
              "      <td>0.32340</td>\n",
              "      <td>0.19832</td>\n",
              "      <td>1.6278</td>\n",
              "      <td>11.2470</td>\n",
              "      <td>0.43489</td>\n",
              "      <td>0.122310</td>\n",
              "      <td>0.293040</td>\n",
              "      <td>0.096680</td>\n",
              "      <td>0.122310</td>\n",
              "      <td>141.62</td>\n",
              "      <td>84.574</td>\n",
              "      <td>0.73919</td>\n",
              "      <td>0.95787</td>\n",
              "      <td>65.936</td>\n",
              "      <td>0.188110</td>\n",
              "      <td>0.116010</td>\n",
              "      <td>1.2959</td>\n",
              "      <td>0.56511</td>\n",
              "      <td>0.40285</td>\n",
              "      <td>1.8839</td>\n",
              "      <td>2.1184</td>\n",
              "      <td>3186.6</td>\n",
              "      <td>0.134850</td>\n",
              "      <td>0.48431</td>\n",
              "      <td>0.86515</td>\n",
              "      <td>0.124440</td>\n",
              "      <td>6.3985</td>\n",
              "      <td>4.3158</td>\n",
              "      <td>127.210</td>\n",
              "      <td>2.8692</td>\n",
              "      <td>7.8980</td>\n",
              "      <td>b'0'</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      Attr1    Attr2    Attr3   Attr4  ...   Attr62  Attr63   Attr64  class\n",
              "0  0.200550  0.37951  0.39641  2.0472  ...   82.658  4.4158   7.4277   b'0'\n",
              "1  0.209120  0.49988  0.47225  1.9447  ...  107.350  3.4000  60.9870   b'0'\n",
              "2  0.248660  0.69592  0.26713  1.5548  ...  134.270  2.7185   5.2078   b'0'\n",
              "3  0.081483  0.30734  0.45879  2.4928  ...   86.435  4.2228   5.5497   b'0'\n",
              "4  0.187320  0.61323  0.22960  1.4063  ...  127.210  2.8692   7.8980   b'0'\n",
              "\n",
              "[5 rows x 65 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 81
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "a885f907",
        "outputId": "d8ba7590-71a5-4413-cd4c-9a19eb0921d1"
      },
      "source": [
        "#Checking the Datatype of the features\n",
        "df_train.info()"
      ],
      "id": "a885f907",
      "execution_count": 82,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "Int64Index: 33613 entries, 0 to 5909\n",
            "Data columns (total 65 columns):\n",
            " #   Column  Non-Null Count  Dtype  \n",
            "---  ------  --------------  -----  \n",
            " 0   Attr1   33606 non-null  float64\n",
            " 1   Attr2   33606 non-null  float64\n",
            " 2   Attr3   33606 non-null  float64\n",
            " 3   Attr4   33522 non-null  float64\n",
            " 4   Attr5   33545 non-null  float64\n",
            " 5   Attr6   33606 non-null  float64\n",
            " 6   Attr7   33606 non-null  float64\n",
            " 7   Attr8   33538 non-null  float64\n",
            " 8   Attr9   33604 non-null  float64\n",
            " 9   Attr10  33606 non-null  float64\n",
            " 10  Attr11  33570 non-null  float64\n",
            " 11  Attr12  33522 non-null  float64\n",
            " 12  Attr13  33507 non-null  float64\n",
            " 13  Attr14  33606 non-null  float64\n",
            " 14  Attr15  33585 non-null  float64\n",
            " 15  Attr16  33537 non-null  float64\n",
            " 16  Attr17  33538 non-null  float64\n",
            " 17  Attr18  33606 non-null  float64\n",
            " 18  Attr19  33506 non-null  float64\n",
            " 19  Attr20  33507 non-null  float64\n",
            " 20  Attr21  27917 non-null  float64\n",
            " 21  Attr22  33606 non-null  float64\n",
            " 22  Attr23  33507 non-null  float64\n",
            " 23  Attr24  32902 non-null  float64\n",
            " 24  Attr25  33606 non-null  float64\n",
            " 25  Attr26  33537 non-null  float64\n",
            " 26  Attr27  31490 non-null  float64\n",
            " 27  Attr28  33032 non-null  float64\n",
            " 28  Attr29  33606 non-null  float64\n",
            " 29  Attr30  33507 non-null  float64\n",
            " 30  Attr31  33507 non-null  float64\n",
            " 31  Attr32  33341 non-null  float64\n",
            " 32  Attr33  33522 non-null  float64\n",
            " 33  Attr34  33538 non-null  float64\n",
            " 34  Attr35  33606 non-null  float64\n",
            " 35  Attr36  33606 non-null  float64\n",
            " 36  Attr37  19071 non-null  float64\n",
            " 37  Attr38  33606 non-null  float64\n",
            " 38  Attr39  33507 non-null  float64\n",
            " 39  Attr40  33522 non-null  float64\n",
            " 40  Attr41  33046 non-null  float64\n",
            " 41  Attr42  33507 non-null  float64\n",
            " 42  Attr43  33507 non-null  float64\n",
            " 43  Attr44  33507 non-null  float64\n",
            " 44  Attr45  32079 non-null  float64\n",
            " 45  Attr46  33521 non-null  float64\n",
            " 46  Attr47  33389 non-null  float64\n",
            " 47  Attr48  33605 non-null  float64\n",
            " 48  Attr49  33507 non-null  float64\n",
            " 49  Attr50  33538 non-null  float64\n",
            " 50  Attr51  33606 non-null  float64\n",
            " 51  Attr52  33388 non-null  float64\n",
            " 52  Attr53  33032 non-null  float64\n",
            " 53  Attr54  33032 non-null  float64\n",
            " 54  Attr55  33612 non-null  float64\n",
            " 55  Attr56  33507 non-null  float64\n",
            " 56  Attr57  33607 non-null  float64\n",
            " 57  Attr58  33545 non-null  float64\n",
            " 58  Attr59  33607 non-null  float64\n",
            " 59  Attr60  32075 non-null  float64\n",
            " 60  Attr61  33543 non-null  float64\n",
            " 61  Attr62  33507 non-null  float64\n",
            " 62  Attr63  33522 non-null  float64\n",
            " 63  Attr64  33032 non-null  float64\n",
            " 64  class   33613 non-null  object \n",
            "dtypes: float64(64), object(1)\n",
            "memory usage: 16.9+ MB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ae35a468"
      },
      "source": [
        "Note:\n",
        "\n",
        "We can see that the CLASS feature is of Object-Type which means it is probably in String format. We will first check the unique values in that series then replace them with either 0 or 1"
      ],
      "id": "ae35a468"
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9fde03d7",
        "outputId": "40cf94c0-6cee-4bae-84cd-874525c84b58"
      },
      "source": [
        "#Listing the names of all features\n",
        "data_keys=list(df_train.columns)\n",
        "len(data_keys)"
      ],
      "id": "9fde03d7",
      "execution_count": 83,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "65"
            ]
          },
          "metadata": {},
          "execution_count": 83
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "eda68de1",
        "outputId": "708a71bb-fa83-409a-d836-1fc5f8686819"
      },
      "source": [
        "#Checking the unique values in the Class-Feature\n",
        "df_train['class'].unique()"
      ],
      "id": "eda68de1",
      "execution_count": 84,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([b'0', b'1'], dtype=object)"
            ]
          },
          "metadata": {},
          "execution_count": 84
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uGxwBIVF-DUz",
        "outputId": "6c20768e-b0b4-49ea-905a-387bd0fc785e"
      },
      "source": [
        "df_train['class'].value_counts()"
      ],
      "id": "uGxwBIVF-DUz",
      "execution_count": 85,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "b'0'    32037\n",
              "b'1'     1576\n",
              "Name: class, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 85
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "f6dae21f",
        "outputId": "e2adb763-4206-4e5a-f08c-b0a67828ad00"
      },
      "source": [
        "#Imputing class labels of train and test into Int type\n",
        "from sklearn import preprocessing\n",
        "\n",
        "# label_encoder object knows how to understand word labels.\n",
        "label_encoder = preprocessing.LabelEncoder()\n",
        "\n",
        "# Encode labels in column 'species'.\n",
        "df_train['class']= label_encoder.fit_transform(df_train['class'])\n",
        "df_test['class']= label_encoder.fit_transform(df_test['class'])\n",
        "df_train['class'].unique()"
      ],
      "id": "f6dae21f",
      "execution_count": 86,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0, 1])"
            ]
          },
          "metadata": {},
          "execution_count": 86
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0415f50c",
        "outputId": "6713cf91-cd1c-485a-cb71-39365b570956"
      },
      "source": [
        "df_train['class'].dtype #The datatype for the CLASS-Feature is now converted into INT type."
      ],
      "id": "0415f50c",
      "execution_count": 87,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "dtype('int64')"
            ]
          },
          "metadata": {},
          "execution_count": 87
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "evPSLi9A-LJW",
        "outputId": "3ef3e808-2fa2-435b-ba59-9162f3b6856b"
      },
      "source": [
        "df_train['class'].value_counts()"
      ],
      "id": "evPSLi9A-LJW",
      "execution_count": 88,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0    32037\n",
              "1     1576\n",
              "Name: class, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 88
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 301
        },
        "id": "6f3096ef",
        "outputId": "253e9415-1141-4b2a-872c-6d76fad485f2"
      },
      "source": [
        "plt.title('Distribution of Class Labels in Training Data')\n",
        "df_train['class'].value_counts().plot(kind = 'barh')"
      ],
      "id": "6f3096ef",
      "execution_count": 89,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f078e28a0d0>"
            ]
          },
          "metadata": {},
          "execution_count": 89
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 301
        },
        "id": "afb8db33",
        "outputId": "3f0c8f09-8907-46d1-e802-1b4646907ed8"
      },
      "source": [
        "plt.title('Distribution of Class Labels in Test Data')\n",
        "df_test['class'].value_counts().plot(kind = 'barh')"
      ],
      "id": "afb8db33",
      "execution_count": 90,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f07804af710>"
            ]
          },
          "metadata": {},
          "execution_count": 90
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "caed7806"
      },
      "source": [
        "## Missing Data Analysis "
      ],
      "id": "caed7806"
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "04EOC4Lg-TOf",
        "outputId": "9ec409f0-5d21-447b-f433-56a54a5d277b"
      },
      "source": [
        "total_cells= np.product(df_train.shape)\n",
        "total_missing=df_train.isnull().sum().sum()\n",
        "# percent of data that is missing\n",
        "percent_missing = (total_missing/total_cells) * 100\n",
        "print(percent_missing,'percent')"
      ],
      "id": "04EOC4Lg-TOf",
      "execution_count": 91,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1.489625122148253 percent\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 425
        },
        "id": "eccd0955",
        "outputId": "74d7b9e9-adb9-4fba-c406-a46c9b637558"
      },
      "source": [
        "plt.figure(figsize = (25,7))\n",
        "plt.xticks(rotation=65)\n",
        "plt.yticks(rotation=40)\n",
        "sns.set_theme(style = \"ticks\", palette = \"pastel\")\n",
        "sns.barplot(x = df_train.columns,y = df_train.isna().sum().values)\n",
        "plt.title(\"Missing values in Training Dataset\")"
      ],
      "id": "eccd0955",
      "execution_count": 92,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Missing values in Training Dataset')"
            ]
          },
          "metadata": {},
          "execution_count": 92
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABawAAAHICAYAAACidK1iAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzde7ju9Zw38Pc+VETRQUIj5w8PhhiHuS6VmTwyHocxD1KkInKIkchgmg4mjFPxhJLIlJHRyOFBMQ2pxykPwzQzPjlEKEYqUand3uv5476XZ03TYe+97rXu31779bquLuv3/X3v3/dzr/v+Y1/v9fX5LpuZmQkAAAAAAEzb8mkXAAAAAAAAicAaAAAAAICBEFgDAAAAADAIAmsAAAAAAAZBYA0AAAAAwCAIrAEAAAAAGASBNQAAa62qjquqQ+fx+tdU1XsnWdN8VdW+VXXuIq/5maraZzHXnE8NQ6gXAICNw7KZmZlp1wAAwJRV1Q+T3DnJnbv70jnj30zy4CR37+4fTqW4BVZV+ybZv7sfNe1abklV/WbO5eZJrk2yenx9QHd/cPGrWn9V9egk/5Tk6vHQFUm+lOTN3X3eWj7j8CT36u5nLUSNi70OAMDGzg5rAABmXZhkz9mLqnpgRqEoA9Hdt539L8lFSZ44Z+x3YXVVrZxelevs4vH72SLJI5N8J8k5VbXbdMsCAGAaNqR/yAIAsLBOTvLsJP9rfL1Pkr9N8tezE6rqpCQ/6e6/rKptk5yU5FFJ1iT51yS7dveaqnpVkpcm2TLJxUle1N1nzd2lWlV3yygk3zfJ6zIKx4/u7qPGa906yXFJnpTkZ0nen+Sl3b3DDQuvqncnuaq7XzFn7ONJzu7ut1XVXyR5XpLtkvw4yWu7+/Qbec5sTZt09/XjsS8kOaW73zu+fk6SVybZPsnXkjy/u39UVcuSvC3JM5PcKsmPkuzZ3effyDq/e+bsDu8kX0ny3Ix2Gb+ouz9zw9fdlPFO5VMy+uwOSvK5qnppRp/pIzL6d///SfKC7v7JutawjnPvnuQDSXZK8tUkneR2t7QzubtnkvwkyV9V1dZJ/ibJH4yf+fYkf5bkdkm+m+Rl3X1OVT0uyWuSLKuqP03y/e5+UFXtl+SQJDsk+UWSv+nu48fPurnv7Z3Hv8Ndkvwmo+/jO25qnbX4aAAAWEd2WAMAMOsrSbasqvtV1Yokz8goBL0pB2cUMN4hyR0zCvRmqqqSHJjkYd29RZLdk/zwZp7zqCSVZLeMwsr7jccPS3K3JPdI8t+T3Fzg+aEke4xD41TVVkkem+TU8f3vJ9k5o8DziCSnVNWdbuZ5N6qqnpzR+/yzjN73OeO1M15vlyT3Ga/z9CS/XMtHPyKjYHfbJG9KcuLse1kH2yfZOsmOSZ6f0b/13z++vmuSa5IcO6Eabm7u32UU5G+T5PAke6/j+0iSjyZ5SFXdZnx9XkatabYeP/8jVXWr7j4jyeuTfHi8y3w2RP6PJE/I6A8m+yU5uqoeMr53U9/b5Uk+meRbSe6S0ffxZVW1+82sAwDAhNlhDQDAXLO7rM9O8u9Jfnozc1cluVOSHbv7exmFt6mq1Uk2S/LfquoXa9H7+ojuvibJt6rqW0keNF776Ule2N2XJ7m8qt6RUQB6Y85JMpNRKP3FJE9N8uXuvjhJuvsjc+Z+uKpeneThST5+C7Xd0AuSvKG7/338Xl+f5DVVtWNGv48tktw3yddm56ylH3X3CeNnfiDJuzIKU3+2Ds9Yk+Sw7r52fH1Nkn+YvVlVRyX5/IRquNG5VbVpkocl2a27r0tyblV9Yh3ew6yLkyxLcvuMds7P/cPJW6vqLzP6I8e3buzF3f2pOZdnV9VnM/pufCM3/b19eJI7dPeR49f9oKpOyOgPN2eux3sAAGA9CKwBAJjr5IwC37tn1A7k5rw5owD5s6NN1XlPd7+xu79XVS8b37t/VZ2Z5OWz4fGNmBuIXp3ktuOf75xR+45Zc3/+T7p7pqpOzagH9xeT7JU5u8Or6tlJXp7Rju2M19j2Ft7fjdkxydur6q1zxpYluUt3/1NVHZvknUl2rKqPJnlFd1+5Fs/93e+gu68e/z5ve9PTb9Qvuvu3sxdVtXmSo5M8LslW4+EtqmpFd6++kdevSw03NXfbJJd199Vz5v44ye+t43u5S0Z/gLhi/F5ekVH7kTuPx7fMzXx+VfUnGe3Qv09GO803T/Iv49s3+r3N6LO9c1VdMedRKzIOtAEAWBxaggAA8Dvd/aOMejg/PqO2DDc399fdfXB33yOjPtMvnz0or7v/rrsflVEIOJNRP+J1dUlGPYhn3VLo+aEkTx3vdn5ExruLx9cnZNSmZJvuvn2S8zMKmm/oqvH/zj1scvs5P/84yQHdffs5/926u7+UJN39ju5+aJL/llFY+sq1eJ+TMnOD64Mz2oX8iO7eMqN2JcmNv+9JuSTJ1uOwfNa6htVJ8pQk3+juq6pq54z6UT89yVbjz+9X+f/v4z+976raLKPP/i1J7jie/+nZ+Tfzvf1xkgtv8Nlu0d2Pv7F1AABYGHZYAwBwQ8/NKBi8qqpu8t+LVfWEJN/JqD/0r5KsTrJm3MP6Lhkd8vfbjFpTrFiPOv4+yaur6ryMAuQDb25yd3+zqi5N8t4kZ3b37E7Z22QUNv5iXPd+SR5wE8/4RVX9NMmzqur4jA6evOecKccleV1V/XN3/2tV3S7JY7v7I1X1sIw2hHwjo+D7txm16ZiWLTL63V8xPsTwsIVecHz45NeTHD5u2/HQJE/MqDf0zRr3wL5zRgc67p9RmJyM3sf1GX1+K8cHaG4556U/T/Lfq2p5d69JsmlGLWl+keT68W7rx2b0R4qb/N5m1Hf71+MDQ9+R5Lok90ty6+4+70bWAQBgAdhhDQDAf9Ld3+/ur6/F1Hsn+cckv0ny5STv6u7PZxQWvjHJpRm1jtguyavXo5QjMzoc78LxOqclufZmXzE6kO8x4/9NknT3vyV567jGnyd5YEZh+k15XkY7o3+Z5P5JvjTnWadntFv81Kq6MqMQ9E/Gt7fMaCf35Ul+NH79m2/5bS6YY5LcOqPP4StJzlikdZ+Z5A8zev9/neTDufnP7c5V9ZuMvkfnZfT5PLq7Pzu+f2ZGtV+Q0e/1t/nP7WFm+5P/sqq+0d2/TvLSjP7gcXlG7WHm9tG+0e/tuE3KEzI63PHCjH5v783oAM3/ss7a/SoAAFhXy2Zm/D/bAAAYvqp6YZJndPeu066FtVdVH07yne5e8B3eAABs+LQEAQBgkKrqTknukdEu2Htn1JP52KkWxS0at0a5LKNdyo9N8uSMdtwDAMAtElgDADBUmyY5Psndk1yR5NQk75pqRayN7TM6sHObjFq6vLC7vzndkgAA2FBoCQIAAAAAwCA4dBEAAAAAgEFYEi1BqmqzJA9LckmS1VMuBwAAAACAG7ciyZ2SnNfd197w5pIIrDMKq8+ZdhEAAAAAAKyVnZOce8PBpRJYX5IkH/zgB7P99ttPuxYAAAAAAG7Ez372szzzmc9MxpnuDS2VwHp1kmy//fbZYYcdpl0LAAAAAAA370ZbOzt0EQAAAACAQRBYAwAAAAAwCAJrAAAAAAAGQWANAAAAAMAgCKwBAAAAABgEgTUAAAAAAIMgsAYAAAAAYBAE1gAAAAAADILAGgAAAACAQRBYAwAAAAAwCAJrAAAAAAAGQWANAAAAAMAgCKwBAAAAABgEgTUAAAAAAIMgsAYAAAAAYBAE1gAAwJJy3erVg3gGAADrbuW0CwAAAJikTVesyJNO+9/zesYnnvqECVUDAMC6sMMaAAAAAIBBEFgDAAAAADAIAmsAAAAAAAZBYA0AAAAAwCAIrAEAAAAAGASBNQAAAAAAgyCwBgAAAABgEATWAAAAAAAMgsAaAAAAAIBBEFgDAAAAADAIK6e5eFXtmeSKJJd391emWQsAAAAAANM1tR3WVfXeJAcleXyS91bVwdOqBQAAAACA6ZtKYF1V906yQ3c/vLtfkmTfJIdX1bPG97UqAQAAAADYyCxaMFxVy+Zcbp/k92YvuvvrSfZJcnxV3au719xgPgAAAAAAS9xi7mTeZvaH7j4nyUVVddScsY8meXeSI6pqWXfPLGJtAAAAAABM2aIE1lX1oiSfrKpjq+qw8fBbkzygqp4zZ+pZSS4RVgMAAAAAbHxWLvQCVbV/kgOSPC/J7ZK8r6o2S3JokrskeWFVbZrk+CTPTnJVVS3v7jULXRsAAAAAAMOx4IF1kq2THN3dX0uSqtolydeTdHd/oKouS3Jkkicn+VV3778INQEAAAAAMDCL0RJkJsmBsxfdfWGSvZIcW1X36e5PJtklybO6+xmLUA8AAAAAAAM08cC6qh5RVXtX1a7joaOT/OAGByyemeTUJA8dD/2mu3856VoAAAAAANhwTDSwrqrnJzkpyc5JXltV/9Dd1yc5Mcn9q+qVc6ZvldHu6zhkEQAAAACAifWwHh+c+CdJntHd36qqWyf5RlUd290HVtXyJMdU1T0zCqvvkOSzk1ofAAAAAIAN2yR3WN8myX2S3CpJuvuaJLsmeUZVPb+7P5PkcUm+lOS87t61uy+b4PoAAAAAAGzA5hVYV9U2VbVdknT35UlOS3LoeHd1uvs/Mjpg8cVVtUN3X9jdf9vdb5lv4QAAAAAALC3rHVhX1dZJ/jnJcVW1zXj475L8IsmRc6Z+M8n5Sa5c37UAAAAAAFj65rPDenmSHyV5QJLXVNUm3d1JTkly76p6z7hv9V5J7pVks6paNu+KAQAAAABYktYrsK6qZd19aZLPJXlSkick2Wd8+0dJDk5y/ySnJtk7ydO6+xfdPTP/kgEAAAAAWIrWK7CeEzw/KsnqJE9L8ldVdVGSXbr7+0keneQ5SR7W3RdNoFYAAAAAAJawlWszqaq2TfLmJG/s7q6qFd29Osm3kvwmyWXjZ61Icm6SdPeqJKsWpGoAAAAAAJacW9xhXVU7JvlMkj2SvKGqth6H1UlyaZJPJTkvyXOTnJjk1Kq6/QLVCwAAAADAErU2LUFWJXl/d2+eZOskr6+qW4/v/VuSizPqUf2Z7n5Nkn27+4qFKRcAAAAAgKXqFgPr7r44ySnjyz2T7JLkoPG9TyR5anf/36rafDz27QWqFQAAAACAJWytDl3s7iurall3X5JkvyTPq6onVtWyJC8Zz7l6AesEAAAAAGCJW6vAOkm6e2Z82OJXkxyQ5Pgk30vy+wtVHAAAAAAAG4+1DqyTZM5hiz9Osk2Sd3T33hOvCgAAAACAjc7KdX1BVW2S5OVJ9ujuj02+JAAAAAAANkbrtMM6Sbp7VZIXCqsBAAAAAJikdQ6sk6S7r590IQAAAAAAbNzWK7AGAAAAAIBJE1gDAAAAADAIAmsAAAAAAAZBYA0AAAAAwCAIrAEAAAAAGASBNQAAAAAAgyCwBgAAAABgEATWAAAAAAAMgsAaAAAAAIBBEFgDAAAAADAIAmsAAAAAAAZBYA0AAAAAwCAIrAEAAAAAGASBNQAAAAAAgyCwBgAAAABgEATWAAAAAAAMgsAaAAAAAIBBEFgDAAAAADAIAmsAAAAAAAZBYA0AAAAAwCAIrAEAAAAAGASBNQAAAAAAg7BymotX1Z5JrkhyeXd/ZZq1AAAAAAAwXVPbYV1V701yUJLHJ3lvVR08rVoAAAAAAJi+qQTWVXXvJDt098O7+yVJ9k1yeFU9a3xfqxIAAAAAgI3MogXDVbVszuX2SX5v9qK7v55knyTHV9W9unvNDeYDAAAAALDELeZO5m1mf+juc5JcVFVHzRn7aJJ3JzmiqpZ198wi1gYAAAAAwJQtSmBdVS9K8smqOraqDhsPvzXJA6rqOXOmnpXkEmE1AAAAAMDGZ+VCL1BV+yc5IMnzktwuyfuqarMkhya5S5IXVtWmSY5P8uwkV1XV8u5es9C1AQAAAAAwHAseWCfZOsnR3f21JKmqXZJ8PUl39weq6rIkRyZ5cpJfdff+i1ATAAAAAAADsxgtQWaSHDh70d0XJtkrybFVdZ/u/mSSXZI8q7ufsQj1AAAAAAAwQBMPrKvqEVW1d1XtOh46OskPbnDA4plJTk3y0PHQb7r7l5OuBQAAAACADcdEA+uqen6Sk5LsnOS1VfUP3X19khOT3L+qXjln+lYZ7b6OQxYBAAAAAJhYD+vxwYl/kuQZ3f2tqrp1km9U1bHdfWBVLU9yTFXdM6Ow+g5JPjup9QEAAAAA2LBNZIf1OIy+TZL7JLlVknT3NUl2TfKMqnp+d38myeOSfCnJed29a3dfNon1AQAAAADY8M0rsB7vok6SFd19eZKPJTl0dry7/yOjAxZfXFU7dPeF3f233f2WeVUNAAAAAMCSs96BdVU9Msknquo9Sd5cVcuSHJ/k0iRHzpn6zSTnJ7lyPoUCAAAAALC0rVdgXVV7JHlfRgH1p5LskGS/7r4oySlJ7lNVx41bheyV5F5JNhuH2gAAAAAA8F+s86GLVXW7JE9K8oru/nRVbZJktyR3HE85O8mFST6Q5NQk90jyZ939i8mUDAAAAADAUrTOO6y7+1dJ/irJOVW1srtXJflJkruO76/q7u8n+eMkz0nysPHOawAAAAAAuElrtcO6qrZN8uYkb+yR799gyu2SXDGee9ckm3f3d5JcN8liAQAAAABYum5xh3VV7ZjkM0n2SPLGqtp6zr1Nxj9uluQrVfWgJJ9Icr8FqBUAAAAAgCVsbVqCrEry/u7ePMlWSV5fVbca31s95zmHJTkxySHdffrEKwUAAAAAYEm7xcC6uy9Ocsr4cs8kuyQ5qKo27e414/F7Jtk2o8MVP7sglQIAAAAAsKStVQ/r7r6yqpZ19yVVtV+SU5P8S1WdkeSlSV6R5KLuvnYBawUAAAAAYAlbm5YgSZLunqmqFd391SQHJHlPkguS7NTd3xVWAwAAAAAwH2sdWCdJd8/2rP5xkm2SvL279554VQAAAAAAbHTWqiXIXFW1SZKXJ9mjuz82+ZIAAAAAANgYrdMO6yTp7lVJXiisBgAAAABgktY5sE6S7r5+0oUAAAAAALBxW6/AGgAAAAAAJk1gDQAAAADAIAisAQAAAAAYBIE1AAAAAACDILAGAAAAAGAQBNYAAAAAAAyCwBoAAAAAgEEQWAMAAAAAMAgCawAAAAAABkFgDQAAAADAIAisAQAAAAAYBIE1AAAAAACDILAGAAAAAGAQBNYAAAAAAAyCwBoAAAAAgEEQWAMAAAAAMAgCawAAAAAABkFgDQAAAADAIAisAQAAAAAYBIE1AAAAAACDILAGAAAAAGAQBhVYV9WyadcAAAAAAMB0rJzm4lX13CS/TfLr7v5Ed89Msx4AAAAAAKZnajusq+qdSQ5IsmOSt1fVEVW1ybTqAQAAAABguqYSWFfVnZLcK8lu3f36JLsn+Z9JXjCNegAAAAAAmL5FD6zHfaq3SbJlktVJ0t0XJDkoyUFV9cjxvEH11wYAAAAAYGEtWihcVXeqqk26e6a7z09yRZITZu939+eSvC/Jm8bXaxarNgAAAAAApm9RAuuqOjDJGUlOqqrjxsMvTnLnqnr1nKkfTnLBYtQEAAAAAMCwrFzoBapqzyTPT7JHRi1ATq+qtyd5XZIjkpxSVUnyoSSHJdmsqlZ09+qFrg0AAAAAgOFYjB3Wy5N8qLv/fdyr+rFJHp5kn+7+QkZB9sOTHJNkdXc/TVgNAAAAALDxWYzAetMk+89edPdPkxyc5LVV9Qfd/X8yCq2f2d37LEI9AAAAAAAM0MQD66rauapeMm4Fku5+f5J/rqqTZ+d095eSnJTkMeOhVd191aRrAQAAAABgwzHRwLqq9k1yQpI7Jtm/qj5bVbdPcmiS7avqmDnTb5/kV0nS3TOTrAMAAAAAgA3PpA9d3C3JQd39mSSpqs8leUOSVyc5MMmnxwH2tkm2TPKqCa8PAAAAAMAGaiKBdVUtT3LrJNskWTbn1p5Jzkzyou5+fVXtnOT+Se7R3cdPYm0AAAAAAJaGeQXWVbVFkuuSzHT3VVV1dpKjq+qL3f2b7r60ql6Q5PSq+vvu/l6SiydQNwAAAAAAS8x697Cuqj9Mcm5GPas/XlVbdfffjMfeNzuvu89L8vkk186zVgAAAAAAlrD12mFdVfdOclKS1yb5QpJ3JPlYVe2f5KgkJ1bVR5O8MsmTkuyUZM0E6gUAAAAAYIla3x3WK5J8ubtP6+5Lu3uvjFp9HNrdP0jyhCTXJ/nLJE9J8vju/ulEKgYAAAAAYEla38B6VZLHVdUj5oztm+ThVfUX3X1VRgcu7p9k1+6+aH5lAgAAAACw1K1VYF1V21bV+6uqkqS7v5/kbUmOrao7jseuzagFyEOrarOMDmJc3d0zC1Q7AAAAAABLyC0G1lW1Y5LPJNkjyRuqatvxrfck+XqSD1fVsvHYtkmu6e5ru1vPagAAAAAA1tra7LBeleT93b15kq2THFVVm3X3FUleleTXST5fVe8ZX5+8YNUCAAAAALBk3WJg3d0XJzllfLlnkp2THFxVm3b3ld39xCSHJflakv/R3Z9bsGoBAAAAAFiyVq7NpO6+sqqWdfclVbVfklOTfLuqzkzyku5+W5KzF7JQAAAAAACWtrU6dDFJunumqlZ091eTHJBRD+vvJNlpoYoDAAAAAGDjsdaBdZJ09+rxjz9Osk2Sd3T33hOvCgAAAACAjc5atQSZq6o2SfLyJHt098cmXxIAAAAAABujddphnSTdvSrJC4XVAAAAAABM0joH1knS3ddPuhAAAAAAADZu6xVYAwAAAADApAmsAQAAAAAYBIE1AAAAAACDILAGAAAAAGAQBNYAAAAAAAyCwBoAAAAAgEEQWAMAAAAAMAgCawAAAAAABkFgDQAAAADAIAisAQAAAAAYBIE1AAAAAACDILAGAAAAAGAQBNYAAAAAAAyCwBoAAAAAgEEQWAMAAAAAMAgCawAAAAAABkFgDQAAAADAIAisAQAAAAAYBIE1AAAAAACDILAGAAAAAGAQBhNYV9VgagEAAAAAYPGtnNbCVbV7kickuSjJad194bRqAQAAAABg+qayq7mq9k1yTJJOsmOSz1fVg6ZRCwAAAAAAwzCtNhyV5IjuPra7D0zy90neUlX3nFI9AAAAAABM2aIG1lW1vKqWJblbknvPjnf3IUkuS/KmxawHAAAAAIDhWJTAuqq2q6odkmzX3TNJjkvy4qr6wznTnpXkgVW122LUBAAAAADAsCx4YF1Vf5Tki0nenOTMqtqtu89O8pYkR1fV9knS3auSnJvRTmsAAAAAADYyCxpYj8Poo5Mc0t17JjkhyfFV9cdJTkzytSTnVNXuVXVYkkckuWIhawIAAAAAYJgWeof18iQXJPlUknT3sRkF1Ucm2bS7X5rkg0l2S/KQJLt194ULXBMAAAAAAAO0coGf//MkD0zy50neliTd/YaqeliSdyZ5ancfvsA1AAAAAACwAZjoDuuqulNVnVBVL6uqJ3f36iQvSfKcqnrMnKmvSLKiqrasqmWTrAEAAAAAgA3TxALrqrp/krMyOjRxJslJVbV3ki8neXdGvasfPJ7+yCRbJFnV3TOTqgEAAAAAgA3XJFuCVJJ/7O5XJUlV/TDJ4Uku7e53VtW2SU6uqvOT7JTkgO6+ZoLrAwAAAACwAZt3YF1Vt0/ym/GzHjA73t0fr6q7Jzmmqs7p7iOq6vQkWyb5QXdfPN+1AQAAAABYOubVEqSqPpDkpIxagXwyyVZV9crZ+919TJILkrx6fP3t7j5XWA0AAAAAwA2tV2BdVbeuqn9MsirJ3hntrn5GkkOSPK2q/mzO9LOS/GS+hQIAAAAAsLStb0uQeyT5eXfvnyRV9a9JbptRH+u3JDmxqrZJcnWSF2e8wxoAAAAAAG7K+gbW1yfZs6rOSfKoJA9N8tWMgum3Z7TbeveMAuyXdPcZE6gVAAAAAIAlbL1agnR3J9kryRYZ7bZ+SHefmOTxSY5K8t3ufnmSpwirAQAAAABYG+t96GJ3n5rk9CRXdvc147ELkpyTZGZ8fd0kigQAAAAAYOlb78A6Sbr7e0muqaqPVdVOVXVWkgu7+8LJlAcAAAAAwMZiXoH12J8nWZbkkCSf6u4DJ/BMAAAAAAA2Mut76OLvdPdFVfWUJJvNtgYBAAAAAIB1Ne/AOkm6e00SYTUAAAAAAOttEi1BAAAAAABg3gTWAAAAAAAMgsAaAAAAAIBBEFgDAAAAADAIAmsAAAAAAAZBYA0AAAAAwCAIrAEAAAAAGASBNQAAAAAAgyCwBgAAAABgEATWAAAAAAAMgsAaAAAAAIBBEFgDAAAAADAIAmsAAAAAAAZBYA0AAAAAwCAIrAEAAAAAGASBNQAAAAAAgyCwBgAAAABgEATWAAAAAAAMgsAaAAAAAIBBEFgDAAAAADAIAmsAAAAAAAZBYA0AAAAAwCAIrAEAAAAAGASBNQAAAAAAgyCwBgAAAABgEATWAAAAAAAMwmAC66oaTC0AAAAAACy+ldNauKp2T/KEJBclOa27L5xWLQAAAAAATN9UdjVX1b5JjknSSXZM8vmqetA0agEAAAAAYBim1YajkhzR3cd294FJ/j7JW6rqnlOqBwAAAACAKVvUwLqqllfVsiR3S3Lv2fHuPiTJZUnetJj1AAAAAAAwHIsSWFfVdlW1Q5LtunsmyXFJXlxVfzhn2rOSPLCqdluMmgAAAAAAGJYFD6yr6o+SfDHJm5OcWVW7dffZSd6S5Oiq2j5JuntVknMz2mkNAAAAAMBGZkED63EYfXSSQ7p7zyQnJDm+qv44yYlJvpbknKravaoOS/KIJFcsZE0AAAAAAAzTQu+wXp7kgiSfSpLuPjajoPrIJJt290uTfDDJbkkekmS37r5wgWsCAAAAAGCAVi7w83+e5IFJ/jzJ25Kku99QVQ9L8s4kT+3uwxe4BgAAAAAANgAT3WFdVXeqqhOq6mVV9eTuXp3kJUmeU1WPmTP1FUlWVNWWVbVskjUAAAAAALBhmlhgXVX3T3JWRocmziQ5qar2TvLlJO/OqHf1g8fTH5lkiySruntmUjUAwIZk1errBvEMAAAAGIpJtgS5b5J/7O5XJUlV/TDJ4Uku7e53VtW2SWWkX68AACAASURBVE6uqvOT7JTkgO6+ZoLrA8AGZZMVm+a1H3ncvJ5x1NPOmFA1AAAAMH3zDqyr6slJPpdkqyT3mR3v7o9X1d2THFNV53T3EVV1epItk/yguy+e79oAAAAAACwd82oJUlUnJ3lOks2TnJzk7lX1itn73X1MkguSvHp8/e3uPldYDQAAAADADa3XDuuqWp7k6CTp7ifPGX9xkqOq6gfd/dHx8FlJrp1voQAAAAAALG3rFVh395qq2iTJO5Kkqp6XZLMkt0vyliQnVtU2Sa5O8uKMd1gDAAAAAMBNWa+WIFV1myR3S7JVVe2VZN8klyfZL8l2SZ6e5H5J9kryku4+bRLFAgAAAACwdK1XYN3dVyX5VJLjk/xRksd39weTPC7JUUku6O6XJ3lKd58xqWIBAAAAAFi65nPo4nuTnJFkjyTXJUl3fy/JF5JcP76+bp71AQAAAACwkVjvwLq7r82oN/W3knykqv60qr6Q5Kfd/dMJ1QcAAAAAwEZiPjus091XdPfOST6X5L5JPtzdL55IZQAAAAAAbFRWTuIh3f32STwHAAAAAICN17x2WAMAAAAAwKQIrAEAAAAAGASBNQAAAAAAgyCwBgAAAABgEATWAAAAAAAMgsAaAAAAAIBBEFgDAAAAADAIAmsAAAAAAAZBYA0AAAAAwCAIrAEAAAAAGASBNQAAAAAAgyCwBgAAAABgEATWAAAAAAAMgsAaAAAAAIBBEFgDAAAAADAIAmsAAAAAAAZBYA0AAAAAwCAIrAEAAAAAGASBNQAAAAAAgyCwBgAAAABgEATWAAAAAAAMgsAaAAAAAIBBEFgDAAAAADAIAmsAAAAAAAZBYA0AAAAAwCAIrAEAAAAAGASBNQAAAAAAgzCYwLqqBlMLAAAAAACLb+W0Fq6q3ZI8Osn5Sc7u7p9V1fLuXjOtmgAAAAAAmJ6p7Gquqhckeef48nFJvlRVW3f3mqpaNo2aAAAAAACYrmm14Xhwkv26+9Du3i/JOUk+W1UruntGaA0AAAAAsPFZ1MC6qpaPe1U/KMlDZ8e7e58k1yY5eXw9s5h1AQAAAAAwfYsSWFfVNlW1VZItxz2q35pkr6p64Jxpf5rkvlW102LUBAAAAADAsCx4YD0+XPGsJMclOaWqfr+7T0tydpLXVNU246mXJvlukssXuiYAAAAAAIZnQQPrqrprRocr/kWSg5J8Nck/VdU9khyf5NdJPl5VleSQJPdKcv1C1gQAAAAAwDCtXODnb5nku919xvj6dVV1hyQf7+4HJnl+Vb07yauS7Jjkf3T3zxa4JgAAAAAABmihW4J0Rn2p9/7dQPdLk/yyqt41vn5hkud2927CagAAAACAjddEd1hX1fZJDk7yf5Nc3N1frKrXZ3TAYnf318ZTX5fkaVW1Msnq7p6ZZB0AAAAAAGx4JrbDuqoenOSLSTZP8qAk/1BVj0nykYx6V7+uqu4znn6vJNslibAaAAAAAIBksjus/yDJ33b3XydJVX0nyYeT7JLkTUmOTPK5qjojyWOS7NPdDlgEAAAAACDJBHZYV9Vtxz/eLcnDZ8e7+wNJjk/yye6+urtfkWSfJB9Psmt3nzvftQEAAAAAWDrmFVhX1fuSnFZVmyZ5Z5KtqurZs/e7+zVJvltVLxhff6G7P93dP5nPugAAAAAALD3rHVhX1W2S3DHJTkkOTXJNkvcneWpVPXrO1H9Jcuk8agQAAAAAYCOw3oF1d1+V5Mwkz02yR5KnJvlQkn9O8r+q6klVtXuSJyS5egK1AgAAAACwhM330MWdk3wrybOTnJJk9yQXJHl3kr2TbJfk4O7+9DzXAQAAAABgiVuvHdZVNfu6c5PMdPdXklye5IlJzu7udyV5epLdu/tTE6kUAAAAAIAlbb12WHf3mvGPVyc5sKoOTXJ+ki8keUdV/XF3X5zktxOpEgAAAACAJW++LUHOS3JgktO7+/AkqaoLxmE1AAAAAACstfkG1t9Osmd3/9vsQHefMM9nAgAAAACwEVqvHtazunvNbFg9p681AAAAAACss4mFzHP6WgMAAAAAwDqzKxoAAAAAgEEQWAMAAAAAMAgCawAAAAAABkFgDQAAAADAIAisAQAAAAAYBIE1AAAAAACDILAGAAAAAGAQBNYAAAAAAAyCwBoAAAAAgEEQWAMAAAAAMAgCawAAAAAABkFgDQAAAADAIAisAQAAAAAYBIE1AAAAAACDILAGAAAAAGAQBNYAAAAAAAyCwBoAAAAAgEEQWAMAAAAAMAgCawAAAAAABkFgDQAAAADAIAisAQAAAAAYBIE1AAAAAACDILAGAAAAAGAQBNYAAAAAAAyCwBoAAAAAgEEQWAMAAAAAMAiDCqyralD1AAAspOtWXz+IZwAAAAzFymkuXlW7JXl0kvOTnN3dP6uq5d29Zpp1AQAshk1XrMzjT//reT3j00/5ywlVAwAAMH1T29FcVS9I8s7x5eOSfKmqtu7uNVW1bFp1AQAAAAAwHdNswfHgJPt196HdvV+Sc5J8tqpWdPeM0BoAAAAAYOOy6IF1VS0f96p+UJKHzo539z5Jrk1y8vh6ZrFrAwAAAABgehYtsK6qbapqqyRbjntUvzXJXlX1wDnT/jTJfatqp8WqCwAAAACAYViUwHp8uOJZSY5LckpV/X53n5bk7CSvqaptxlMvTfLdJJcvRl0AAAAAAAzHggfWVXXXjA5X/IskByX5apJ/qqp7JDk+ya+TfLyqKskhSe6V5PqFrgsAAAAAgGFZuQhrbJnku919xvj6dVV1h/y/9s47zK6qauO/yUwKLUgLCQqIwiyagqKIH0UEESlSpPjR4RMREEKTWFBRkY4NKSK9qpBIEQERAUFEUUBBkUUVpIYSQjfJZL4/1r7kMpkkc1rmnrnv73nyZObemXfec+7ZZ+2z9t5rw5Xu/j5gHzM7HfgysDywhbs/Mx98CSGEEEIIIYQQQgghhGgh5kdJECfqUu/21gvu44EXzOy09P1+wOfcfWMlq4UQQgghhBBCCCGEEKI9KX2GtZmNBQ4D7gSecvdbzOwYYoNFd/c70o8eBexgZl1Aj7v3lu1FCCGEEEIIIYQQQgghRH0odYa1ma0J3AIsCKwBTDKzTwCXEbWrjzKz7vTjKwJjAJSsFkIIIYQQQgghhBBCCFH2DOu1gAvc/bsAZnY/8AtgA+AE4DvAb83sOuATwB7urg0WhRBCCCGEEEIIIYQQQpRew3oc8NHGN+5+PnAG8Ct3f93dvwTsAVwJfMzd/1Dy3xdCCCGEEEIIIYQQQghRUwrPsDaz44AZwOPAOcAGZranu58H4O5fM7O1zGxfd/+Ju99c9G8KIYQQQgghhBB1Z3pPL8M7O1pGRwgh5je9M2bS0VVsPm0ZGqK1yJ2wNrPFgInAo8A9wBbAO4ELgB3N7DF3vyn9+L3A8wW9CiGEEEIIIYQQQ4bhnR184/KnCuscte0yJbgRQoj5T0fXMJ794R2FNJY+eO2S3IhWIVfC2syGA8cBf3f3Q9Nrw4BlgGuJ0iAnm9kRwDRgS+DGUhwLIYQQQgghhBBCCCGEGJLkSli7+3QzOwV4penlx4DN3P0FM/sRMBXYDRgDHObu1xR2K4QQQgghhBBCCCGEEGLIUqTAyz/d/d9N348CXgJw92nAhe6+A7Cpu/+6wN8RQgghhBBCCCGEEEII0QbkTli7+0wAM+tML40GHk6vTQS+mX7uzYIehRBCCCGEEEIIIYQQQrQBhbfQdPee9GUHsKaZXQs84e5fLaothBBCCCGEEEIIIYQQon3IVcN6DqwEfArYy93PL1FXCCGEEEIIIYQQQgghRBtQZsJ6InCj6lULIYQQQgghhBBCCCGEyENpCWt3v70sLSGEEEIIIYQQQgghhBDtR+Ea1kIIIYQQQgghhBBCCCFEGShhLYQQQgghhBBCCCGEEKIlUMJaCCGEEEIIIYQQQgghREughLUQQgghhBBCCCGEEEKIlkAJayGEEEIIIYQQQgghhBAtgRLWQgghhBBCCCGEEEIIIVoCJayFEEIIIYQQQgghhBBCtARKWAshhBBCCCGEEEIIIYRoCZSwFkIIIYQQQgghhBBCCNESKGEthBBCCCGEEEIIIYQQoiVQwloIIYQQQgghhBBCCCFES6CEtRBCCCEGjWk901pCQwghhBBCCCFEa9A12AaEEEII0b6M6BzBZlfuXEjj2q0vKcmNEEIIIYQQQojBRjOshRBCCCGEEEIIIYQQQrQESlgLIYQQQgwxpvXMaAkNIYQQQgwdZs7orUSjKl0h2pneGT0toZEXlQQRQgghhBhijOjsYovLTyyk8ettDy/JjRBiqDCtZyYjOovNeSpDQwgxOAzr6uDusyYX0vjA3mP61f3X6c8W0l1lv6UL/b4QQ42Ork4mn3pVIY0xX9yqJDfZUcJaCCGEEEIIIcQ8GdE5jO0m3VFIY9J2a5fkRgghhBBDFQ1tCyGEEEIIIYQQQgghhGgJlLAWQgghhBBCCCGEEEII0RIoYS2EEEIIIYQQQgghhBCiJVDCWgghhBBCCCGEEEIIIURLoIS1EEIIIYQQQgghhBBCiJZACWshhBBCCCHmwbSeGS2lI4QQQgghxFCla7ANCCGEEEII0eqM6Oxiy0nnFta5eru9SnAjhBDznxk9vXR1dgy6hhgYM2f0Mqyr2LkuQ0MIIfKghLUQQgghhBBCCCHmSldnB2f9cnIhjb0/M6YkN2JeDOvq4E/nPVdIY509lyrJjRBCZEMlQYQQQgghhBBCCCGEEEK0BIOesDazs8zsFDM72cwWGGw/dad3Rjl1EcvSEWJu9MyYVolGVbqiHszoKeez66tThm5Z3gaDaSV4L0NDDC2qqgtdhu78qjU9raenJTTE4FHFNTCtZ2YJmsU15tff6atRlve+OtN7egtrlqExEGaU8HfK0BgsekrwXoaGGFr0zijnmihLR8x/emcUj9n9aVSlW2cGrSSImY0ArgSeBS4BJgDXm9lO7v7EYPmqOx1dXTx7+rGFdZbe76tv+753xnQ6uoYX0ixDY7CYOWMaw7pGDLrGUKOzawTXnb15IY1Pfe6afnWvOmezQrpb/d+1hX5/MJnRM42uzmLXWl+NMjTL1JkbXZ0j+N7PNi2sc9hOv5lN9+hfFNM94rO/mfcPtSgjOkew0xWfKqTxs22uK8nN4DCtZzojOovFsTI0hhIjOrvY4penFdb59Wf2n1130pnFNLf7fKHfHygjOjvZcuKFhTSu3n63t30/raeHEZ2dhTTL0BADY0RnJ9tMvLGQxhXbb9RHcxjbTvpDIc3Lt1uv0O8PlBGdw9h+0t8LaUzcbo3ZNHec9EAhTYBLt+t+2/fDOzv4/C8fL6R55meWK/T7A6Wrs4PvX/5MIY1Dtx1bkpv5T2dnB1de9nwhja13WHK213p6euksWIO7DA0xOHR0dfD0CU8X1hk3YVwJbuZO74yZdHQVn6Nals5gUIb3vhodXZ1M/nGxmD3mwI1me62jq5PJpxR7VhpzQLFntVZjMGtYjwCmAPu7++vANmZ2KXC0mX3B3d/MoNUJ8MwzxQLyUOH5qa8U1pj+xOxjBpMvOr6Q5phdv1zo9webBy4q9uDavevsD869PdPpKJi4KENjIMzsmc6wgn+nP40Xpk4vpPlEP9cqwIsV6ZZNVed14qTdC2luv90Fs732s8uLaQLstO3sulXwypRinz/0fw28WlC3P83XKtCsiv++WI3X6S/+txLdKtjr+hML/f65nzx8ttemv/hqIc05n9di/YH5dV6nT3m5sEZ/XqdPmVqB5kuFNKvS7U/zc9deVUjz7M22KvT7rcD0KS8W+v35eW+ZNuWFQr/fn9dpU4ol6+Z0/NOmFKuJ27/XYnWR+9d8tpBm6C4422tvTin2zPnEE7MnT157sdjxh+7ss8pfebHoZzX7SpOpha/V2VdbzZzZy7BhBTcH7EfjxSlTCmk+8UT/KYlbflVMd4NPL1bo97Pw3EtF74Oz99EmTy3/GgB4dmqxe9YiTxTv/w+Eya8Ub689T8yfmbDPnfVQYY2l9l6xBCfz5vmXi92v+sthATx/3l8L6S6554dme+2Fl4u1gWlz8FqF7gsvF7sHzMlrGTTlcPudIdHR2zs4SxHMbAxwC/AFd/99em0h4I/AZe7+3Qxa6wG3VmJUCCGEEEIIIYQQQgghRNms7+6zLQsbtIQ1gJkdBuwBbOzuz6XX1gV+BGzq7gMaXjCzkcCHgaeBoVW0RQghhBBCCCGEEEIIIYYOncA44C/uPttyjsEsCQJwGrAicAHQKDj7JPAoMOB1rOnAihVpE0IIIYQQQgghhBBCCDE/eHhObwzqDGsAM1sSuJioaX0hsDdwh7sfPKjGhBBCCCGEEEIIIYQQQsxXBj1h3SCVBxkNPO/uPx5sP0IIIYQQQgghhBBCCCHmLy2TsBZCCCGEEEIIIYQQQgjR3gwbbANCCCGEEEIIIYQQQgghBChhLYQQQgghhBBCCCGEEKJFUMJaCCGEEEIIIYQQQgghREughLUQQgghhBBCCCGEEEKIlkAJa9GWmFnHYHsYbBrnoA7noiqPVZwDM6vkvlqVbt0o+bOqpA3UqW1VQdXttWTNtm9XFd0H275tVXS9KmbVjDq0gTq1q6qosr1WoNvWbatObaDd25b6g/WiLm2rTu1KOYwS6O3t1b82/Nfd3d3Z3d3dMdg+5uFxWNkeu7u7R3Z3d49IX7f08c+P89FXv5V06v6vu7u7c7A9zMPfO7q7u0f3ea3w9dWdKPscdHd3jyzbaz9/o5Jrt6TzqnZVk3OR4kxnn9cKXQPd3d3Ldnd3L1WH45+D/0piVxnnoo7ns6JzqZhVs5iVdEu/fsvyqrZVj/OgmNWv/6raq/qD5Z7Plj4Xalv9+m/Z57c6ns+KPqOW6g929Pb2DnbOfFAws0537ylJawFgNDASeBKY6e6FT6yZjQPeBSwC3AL0uHuvmXXk1Tez1d39H03fd7n7jCKaSWdtd7/DzIa5+8y8On00dwEeBu5y92lNr3cA5PFrZl8mPp8Ty/DYpLsqsDqwAHCJu08vweviwKLAe4F/uPszJfrdFFgBWBZ4BDivrPaQ9DuAjjKuBTN7P7Aa8CJwq7u/XlQz6W4DvAeYAZzVrJu3PZjZocAf3f1PZXhs0v0O8CPgxTLuLUnzHOBs4M/uPqPPe0XuMb8CJrv750qw2dA8kPisZgJHuPubTe/l/aw2At4NLAc48IuSrtcPJ68LEtfsr0q8J5bWrpLeskTbetDdHy5Jc31gJSIWXl/i9boL8Pfm+FWC5kHA+cDUsnwm3dOAnxL37VLalpn9ErjP3b9eks2G7q5ELJjs7mf0eS+v1zWINrAQ8Czwu5LaVmVxq4K2VXrcUsxq75iVfrf0uFVlzEr6ZfYHS49ZSbf0uFVFzEq6pcctxazKYpb6g+oPqm2V3LaUw6hPf7CRFzSzpYGX3f2NPD7bKmFtZqOANZs/CDPrpECC2cwWAb4HrAfcCZzm7reX4HU0cBnwJjAc2BeYDCzu7k8V0HwcuAeYCJzr7q+Y2eLu/mIjeZ1D9x3A3cAH3X1Keq0LWNDdX87p9R3A34F13P3ptEThI8AUd78/p+aixI2tAzgBOKGkm9GiwI3AXcCSwBeJAYxed/ecmqOJwLkwMAVYFbge+Im7P1DQ74bE8d9FDAh8HBgLnJf0p83xl+eu+wFgJ+LG+UDT67muq/S76wPfJtrAK8Dv3f34PFr96J4I3A68D/gGMJ24+V+dJ6Ck47+TuG6fIM7DlWb2GeAKoCvPuTWzdYn7yhrp+0WBccS18becbXZ94MfuvmaT908BPcDp7v5KVs2ksx5wIXHNPgUc6O6PFhnISl5/SFyf/wOcCXQTbSyXVzP7GHAscS98A9gdeJBIhpxdwOuGwNHAy8A/gOWJ0luXARPzdKjMbBXgk8Dl7v540+uFBl3TdXUEsCIwFfipu59ZMJmyPnH8U4B3Euf1NSLBcnuBe8tHiLb6KyKGTXT335vZJsDvgM7GIGEGzXWBE939f9L3o4iHyoXc/T95fCad9YFT3f396fsVidjVmXznubesS8Ts14HrgK+7+5SiA8RmtgHwfWJAfDXgq8Tn9g5gUk6vGwDfIfouzxNt9g7gYnf/VYGO9IaUHLeqiFnp90uPW4pZ7R2zkm7pcauKmJV0S49bVcSspFt63KoiZiXd0uOWYlalMUv9QfUH1bZKbFvKYdSrP9ikfw5xX/mTmS3QnLgeCK1Zp6Q6jgSuM7OrzWwHAHfvaTQWMxuTQ3MCMbqxEdFwLjGzsY03U+IxD4cAD7j71kSH9FjgDOCnZrZvVrE0WrQI8Gfg98BawPVm9i2i0VPgAW0/4mY2xcyWtxil/CVwRGoAeRhPjMY+nUanLgC+BlxqZsemgYasHAicAnwU2IS4MZXBl4lRs88D9xEPP8cBE83shJS8z8oRRCDeEvgJcaPrBo43s2UK+j0c+L677wucSnQijiECyXYFdE8iOlLnmdnZZvYhiOvKzIalh62sHELcNBs35x3M7D0Q17SZrZbT64HAKe5+CPDPpH0IsAPRpjOROgp3Ew+ppwPXAIeY2Z+Jkd+ZBW7044Fvpb+zbdK/DNifON952IMI7pjZ3sT1OgZYBzg5pybEeT3A3T9IdFJ3BSg4MHQYcJy7/5gIoicAKxPX62mWr97WwUTndF9iwPEq4CbivpD3moK4dn7o7psR19QJxP12M6KTmodTgV2A75nZkemBBXfvSe0qr99DgUvdvZto/3uY2WhPq3hy3mcOJDokWxOd0hOJuPV5YOc8JtPneycwCbiN6PQeaDHD5PTUtjI/+BPt6ofpb2xGzAa9E/iamb0vj9fEvkQbxcx2B34MfBrYnLj35mE88AXgw8RD705QuF1BtIPj3P1QYrbmD4g2+2niYSWv5snu/inivvJPYrB9RzNbqkBiqYq4VUXMgmrilmJWe8csqCZuVRGzoJq4VUXMgpLjVoUxC6qJW4pZ1cQs9QfVH1TbKr9tKYdRk/5gI1dnZp8Glk3J6hWBc83sKjNbaqBabZOwTifts8SHfC2wv5nd1Ej+mtnKwFdyaH4G+J67P+Pu3yISwjum9z8C/F9OrzszqxP+AeAZInF5ObChxczuAePuve7+JDEqtyLRuI8EtgEWSEngd+f0OoFIUJM0xwE3AIsDh+ZMLk8G/pa+PpQY8dmPuEkvB2QKnsnDTsAF7u7EeTjOzA6zAsXq0++uQQRPgA8BfwAOIALV8kRZl6yaqxBLPf/r7jcSI55XEA9U38zrOZ2He5M+7v66uz8P/Bq4Gtgr53XwTmAx4HPAl4gR+lPNbGIK9uOJay2L5nuJJT8Tk9ebgPuBPdOPjCcCaVav7wXM3S9KL20PfDN5Pwv4oJmtkEWzqaNwK3GcE4GtgBHA42Z2l5ltntPrDsRgE8R1dQ6wLbFaYH+L1QhZuY5YpgcR4CekwHcoMNpiGWNWrwas5O6/Ti+dDXzOzCY1OrtZH9TNbGFiQHDh9NJuwBfd/WAiOI8G1s6pOQLA3Z8gSu/8AXiAaF/Ds2gm3VHA00QSBXef7O5/BS4mBjOPtIyDoma2XNI7grj3jyHuWydZLN8cT4o3GXVXAlZ09/OS18uJmS97pB/5IhHbsmqu3NSutiUGCHcj2sV2FkvCMpE6SjOA3xDX7BnEAO7KwCtmdonFjJOsXrcFXk0vTSD6BdsRMyG+nT7PPPyeuO9D3FO+SvQDzgDWSn2NLF5XAd7r7le5+4tE+/+amf3AzBZKP5M5HpjZYsR9pVGqYHtiUHhPosO6VtbOdNJcEHgJwGO57uvECqQZxAN7ZqqIW1XErKRbetxKDziKWW0as/rolha3qohZSbf0uFVFzGrSLTVuVRGzmrxWEbcUs8qPWeoPqj8Ialulti3lMOqVw/BZqz72Bc40s26iH/Qw8G8iZzYg2iZhTXTsjgduJjqlnwd+DmxtZr8jRuqyllowYur9w2Y2Mr12OrBh+vqrpIaakXcRIycPWow+POPuh7n77e5+NlF2IteIn7ufCTwGLOLu1xM1hb5N3FDzaL4b+AtwtJldCqzt7l9x95OJC3QMsx4ysnAn8CUzGw+MdvcT3f0Jj3IrY4l60VlYH7jCU10udz89+duEHJ1ciKCQRgevI25sFwKru/uRyettwNJkmPnSpHkzcJCZbWmxZHVz4BfJ8zJJNzPp5vEzYH0zO9zM1jKz4e7+Rrr5jSXffWEyEYQfcPc/EgMXBxCB5XRimdEFGTU7iFHu5gB8LPDxFJh3JToVWXkpeW0E1G+6+43u/qa7/4Y4v7kGBFJH705iZvzSxMPptsBpxChoVl4jZvBva2ZPAwu4+w3u/pC7f5e4FyyUQ/d2YCUzO4FYrvhk8v9vYkAoTymfscSypMZ1fDMxi+QNopOaeQaAu79KXDfjzewaUk379N6jxL0y0z02aV5GdJi/aWY/ApZw91971INrLIHLhEeN0guALc1sezNb1mKZ5ovufhRxPYycu8psvEa0o1vc/RJixsPF6b0fETMCrszqleiYnmtmnU0JmaOBbVKHcC8iEZKFXuAgADN7F3C8u1/r7s951MLLc/xv4e5nEfVgP0wM3C1APFj8i1iNkoWRxMyMHczsLmCYu1/h7ne5+97AUkSfIQ83Ax8wswOAR8O6v546q+OIjnoWViE+ZwixXxIxawUiMZhrjwSP8l0XA4eb2eXA0+5+W4oFNxHtOdOMiqR5JbCLme1qZt8G3unuVxCza5dL99ysXnuI+PexEuPWc0QH/6ESYxbJx8lAR4lx6+XktQ4x63Vi8sd2ZcWsdB7/FF+WGrPGEQ96ZcesC4GDy4pZTboTiXtWKXGrKWZtXmLMgohbXySWP5cVt0YD51jMdCsrZkHErfFQftwqOWbBrLi1fclx62ZgzRJj1qpUG7O+VHLMugLYueSYVUXbYY8L2QAAHNBJREFUqqJdQTVtq7J2BaW3rVFU065AbavUtpX6gj+n/BzGs0Qfq8wcRi/V5DBepCb9wabjvomYAX8VcKO7H0Fc/wOeYd1WNaz7YrFZYhcxQ2N3d39vxt8fQdwkHnf3N9KNfRHgIqJjuae7b5jDVwdRK6axcd+oFPwao2sXuftaWXWb9DclbnBPEgnmLS3Vsc6hNYwYlXo/EdQe8ZhpjsVGhJd4qjmYQ3sLYFNiNsodRMMfTpQfyTrStwixaeXr6ftGIzqYWF7yVXc/P6fPJYlRspeBTxAlXM4Flkhe18ihuTyxrHQMEYivc/fTLcqj/Czr8fejvzlxE5oBNOpzLQhs4O4bFNFu+huNDT1PAFZz9y1yaCxGFOnvSUFpupmdSTzwTnf3PeYhMSfdfutmmdlewJbunntZURqpHU+0iTvc/UgzG+nu/82pN5wIQJsAj7r779LrewJbuXveAZePEksLNyRmvt1AnNePeSzfyo01bVhhZp8kOhineY4NQlJgNyLBtCdxj72BOL8buvuWOTSXIO5X44i6pb9z97+Z2X7AZu6+VVbNJu19iJHpB4n71nBiY5BN87Qt61OXzqLEUC8xkj4mT7tKOku4+wvp606ig3MO0flf0t13yaOb9PrGsD0pcK026a5CrD5aCbjf3b9lZu9w98wJoHRvWZWog3urxwAuZrYHsI27b1vA53ZEwmtloqbijURfYRN337SA7lvXgsXmOKcRsevUnHpjiE7kA0Tn+ank95PAJ9w9z+yP5YkZFGOJ2SQ3uPsfLFazbeaxPDgXFjUqdyTqFjceSkuLW2XErKTz1v3ezEa4+7Qy4tYc/lZZMesgYtJC4ZiVNMcRDz2PuvsN6bU9KRaz1iIGmzckEsFlx6xhqa9RNGZ1EJMqXqKkmJV0FwT2JiaYPE20rUJxy2KyzT5Em3+IEmJW0l2AuNZnpO8Lxy2LWe/PpX7gMCIxUThmpXPQSMaUErea+qsrEzGrm4IxK+mOJQYxPw78oay4ZWY7EwmvlYl4VShmNd1Ly45ZyxJlHe8lYmwZMaubeBZcjIhZvysjZqV+1ReIEiAPEyuwi/YHhxPPso1zWlZ/cAzwQrr/ldIftKa62o04mL7ek2JxoMOjVMnKxDNMGf3BJYn+4CeA21Lir6z+4GeB/yXabZn9wQ6IBHVJbWscsC6RmNydyBEVbVsrELmMJYgyI7n7g43BlPRc2chhzCQmYULOvmDS7e2bEyjSH2y6RhcFXk3taqS7/7dIX3BOuYum9zP3B/sef+oPHkz0YcrqD76HWLHwkLtfnnJZ53mUYhsQeWrr1o5GwLSom7Ips+ocvZHe35BZo5RZNKeZ2ZtEB/rUdGN+yczuIW70BxTwunxKLJ/mTbuLE6M+l+TUXIm46ZyejnkCEUghglJer6OJC/sgInA2+DoZR6SaNLuJ4HET8XA6ljjuR4Cjcmi+YmYrmdkn3f3Upgb/g/QZ5tmsptERW5CYsX6umf2TWE5yPjGT6Cc5NacTI36nE4mvxrKK48k36+ttuPs1ZnY/EYjGEp3pfzFrqUpm+ibWfFZN9I2JayGPzylNXzfqkjmxlGi9nFYhOmN9g9PCRFv+Rh7B1Mnrdfcnzexl4pw2aovlvtGn437MzM4lNupodFL3Jud5TZ/V7Wb2FGm2NZGs+Bcxiz8XjQCf2ldv8n99ut/k2hnZ3V8jzVBL5+BIIiH2KvmvqxeAk+ztg4HDiQegzJ9/ejgfnbR/alH363+JvQ0WIx6q9smouSCR6BiRPqde4vpqJADWIJbsZfW6ELFcvStdRz2kjYfN7EliRv//5PQ6EngqeWw89C9ErGg6IofXxnkdRZzDB4hR/21IZaGyPpw0ecXdbzOze0n313Q+9iHfeW147XL3SWbmxEDme4gVMpNJM44yaDY+qwWIgcWZjQ6ru1+U7jOZNyFu8trjMdsFi1k130leO0gzT3NovuTuB5vZkh5LNRvndS/yXQNjiQf8Bd39t2b2KLABsbrLiM2s9syoOY6Y6bowsQRyps9aagw5Y1ZKqL2TKLP2R+L8NhJhDxGzazLFrTl5TQ+ro8kRs5o0FyGSXk+a2XNEsqpRwzlzzGrSXYCYkXceMCN5HU6OmNV8/O5+k5lNIWatvQ9Yk/j88+zr0visRhPLtnsayZW8MavpWh1BbCo208waS9YPIyY15LmuGrod7n6ymS3k7q9Z1MAcRY64ZWaru/s/0uf8YzO7jRgM2oQ4J5ljVh/dxvNVI8GWO26Z2fvc/V5PG8434lZKJjxFPMtkill9dBuDS50pydxB9OvzXK+N42/0Vx8m+tifSf9njll9vD4DPGNmdzOrbXWSI241vCZPl5jZP4AtiP5g3pj1fne/p+leOszMekuIWY3z+h/SJJsUs44mf8xqaD5AlCpa2t2fTe2qSMxa293vSPeT08zsD8RgwCeIVRBPkr0/+BF3/zMwI/XVGv2Aov3Bddz9T+4+ufFaU5L5aWJCV9b+YMPrW89YTcnqhcn57NI4r8RAVQ/RH1yCSFzm7Q82jv954JbUrhpeh5O/P9jwirv/IrWtzYkYuwXxfJ+1ba3jUQO4+TmL9DeKtK3G9fo0qXyFmf2CyDc0krRZ21ZD81HgAHv7BJnh5GtbOwEPm9lfUg7jPmL1/ErEeb2XfDmMnYGHzOyupv5a0RzGLmb2EPCXpn5Fo0+Vqy/YR/dtXlMcWJR8OYzGeb3L3ael/uATRKmRXP3BPrm8A9x9PFFnuzFA/L9EWZwB0xYzrG3WSMfxwH/c/RR7+8jUZsBvPcOmg02aJxAzrE9p+oCWIRLWW2bRnIfXDqJB7k7sCJvH6/HAk6nDuwiRvL7c5jFiMwDdE4B/u/tpTa+tRCSVxhc4rw+l5M8o4oHlSWBkjoA0x3OaPq9cO+s26Z4IPJZ0hxODIgsSyb9/ZDm3TZonEUtTfppeH0Y88H2aKG2S9boaBazp7n/q570is39n07VI3M6Et9pXpt1g56aZ9BYjrqtvl+i1g5i19EF3nzQHiQFpNr23pLs/bzl2756D17faqcWo9cfd/ZySvFZ6DZSk2RjtHgcM96Yd0gv6hFnX1rjUacuiuwixAdZ6RN39Uz3KAWFmi7r71Cx6/WjeSQxc3t7nZ97j7o+UqWtRG22Cu3+hDM10n10NWM/dsw7ezU238WCZaffuJs31iUGQUz2W/zXeX42Y+XNsAa93J6+VXQNNcSLPDutzPAcWsyCWAO71tw+UZ9G8O2ne1uhnEQMOK7n7PRm9jibK97xJDIZ8wd0fM7NF3P2VLFpz0BxOJDsnA4s12r41DWSVpWsx0+Ygd/9OiZrLA2t5LAsuovkEkfSa6WmGXY6Y1aw7gphZ+AxRtqLhdSN3P7eA5r7p8x/t7nlKgPSn2+81UFCzcfzPEdftG8BSHnWni+iOIhIozxLn9cn0M2M9EplZNB8H7iFq4Z7j7lNtVvKzK+v134/upKT7ipkt5rEp+wjgXVni1gA0VwMO9NiAK6/XicC5fXQ/BHwoS9yak9f03sLu/mqBttXQ/WXSfbnJ61rEbM3jcmpeDpydNBsz14d5xk22BnBOM8XrAXjtIvZKMmIm4ID7sH00ryA2RnvZUmItPWcsmyNmvYOIgR/0pgk36b3hnmNDwP4007Ev0HR95ekPzlXXYpLfl7K0rQFovg9YN0d/cG7ndZGknaltzcFrBzE4+oqZrU7kcQbcrgbgtbJrIA9z0e0hJocsB/w9Y9vqT3MYMbFvqsXkhkz9waT5d+Cj7v5UenZbBZjSiIN5aNJdJ/VThhFlwV50d08/k6k/2I/XYUSlgCnu7haDNodm6QsOxKtF6Z21M/YH+9Nck8hpvbUCN0fMauRDzwXuc/cTzWxjYtXdKUSN8BeyPL8M+RrWTQ92SxEzyj299dZGgB51lfIkVZdKOg3NYelm9BSwT9bAPDevHrN/7gOOKeC1i0ii4u6veNSqyZtQ6nsOHmp4TZoPEnV1ipzXx5LWmx71D9/w/Mnq/j7/xgzQIsnqpYh25Elrurtf7e6Xuvu9BTQ7iIL0jUTdTI8NGCfm6fARs1KvM7OrzWyH5jc8lqiMzaHZr66793iM/Pea2TIe5XKy1FOao2Z6fzoxE7BMrzOJm/6Ak9Vz0mzgkaxeOuuNfi5em9vp854hWT0Ar1VeA3nqrfenOSP9/zRpY5CSrquehk/Pl6yYQJTW2YiY8XGRxZIqUucszwZjzZoPA5c0fz4WCZtMDycD0fWo8b9/WZrpmr0/68PJXHQb5/VZi0Rw1nthQ/PjxMqVi/tc9/8idjIv4vVBKr4G0vW6aJ7YTf/nYFzSfQT4V5bOeT+ajTYwtnEPIPowmR78E4cQA7dbEzH2GDO7CPiFme2bsf33p/kgMdPlDGJDmH1TRztzsm4euvt7DFpkGgiZh+Y+7v4YkWwpqnkBUVtwb3jbRjl5dR9Iumcmr19IXrOuDuur2fj8f17g8++rO9s1UIJm4/jPIFYG7uw5ktX96N5PbJh+BnBGw6tnS1Z3EDPr/0zMLF+TiItHERNPZuS5/vvR/SBwvZl9i0he4DF7K0uyeq6aqc/8T7LPVOyru1Y/Xv9KXL9FvX7XzP7jUYM8c9vqR/cDwG/6eL2Tprq2OTTXTJpHESV8Znj2ZPVAzmmeZPWcvH6HKEE52d1v9WwJtb6aazDrnN6ZvE7JGbP2I0pBTrFYKX2Qmf3aYmLTKjn0+tUkBi6+bmYfSH7z9AfnpruWuz9ExrY1D8010nPxgNvV3HTTeT0JeC/kilv9eb0yeV3TY/XBgNvVALxWeg2UqHs5MRt2UY9Z0lknM/WneQWxOeQHUi4na9saD1zvkQB+P7GK/QSi3R5rMRiah4bu00n3AmJG/WVJd2SOeNjX6wXEbPLLzOz4FAuy9gXn5fXo1MfI2h/sT/O7xGqDY1L+KXN/0CNZPZaYUHGiRQmXfYma898H3sj6/DLkE9ZNJ+TjRMfv5NRg3lpWU7Lm9PQzmTfUmJtu089kmlUyJ83G+3k7/P3o/qgfr88X1PxhUa9VfP5z0+3rNUuDnNc5zevVYiTys8CBRIH//c3spsbDjkVd2AkV6K5MlEcZ8HkYoNfvZr3RDVD3WyVrrkyUV8jEAL1mHZkdzGsg0zkYgOaqxKZeZV9XeT+rzwDfc/dnPOr3/5mo1YWZrUP2cgVz0twxvf8R0qahJeuuY2aHZumcDEST7A88AzmvHyGWFJah2ez14Jz3lsG4BjId/zx0d0jvf5SUtCzR6zpESZg8XncmNjCESNQ8Q2yONImoY7xwSZo/IR7SNiTHJrYD0P2YxQDDgGdWDUBzY4uBqwEPuM/jnE4kNgNaZKB6GbxulI4/672l1M9/gF43zHoO5qJ5evK6WUXnNbPXNIj0JPF5r0gkwI8kVu+NMrPjzOzdWb3ORXcbokTOsRYrw0rTBI41s/dmTa4OwOtxFjNWB3y9zkVzS6Ks17EVndeG1wEng6u4Bgb4+WfSnIfuVsBIMzu+7OsqHX8mTXirvU4gEokkzXHAb4hSIBNs1qq+opo3ELPLD7WYEVuW12bdEVna1gA0D7eY1Jd10GZu53VR4LAKzuuXktesEw8H6xrIpDkP3d8S5/WbFXg9LM/1SqyA+lv6+lBiVvA+RN9yOWI2eB76092PWCG1HLGPWJmay5rZ8ln6ggPUXcHMlsvSH5yL5j7M2itjXA6fDcYBT1uURduVWHm5HlHeLDNDPmENYLH84GpiqevNwB1mdquZbZn14XQgmlV5rUIz7/HXyetgHH+LeV2UqEV1M3A2cYP/ObC1mf2OqDX5QAW6t+XQHYjXzDW6Bqjrc/ztfJp5jn+wvNbpGri1As28x2/A1e7+sEVdLogkxYbp668QG2+VqfnVHJoD9Zp1490qNAeim+ccDJbXOl0DX67A61fIsUcGUQbsFHd/0GLV0TPufpi73+7uZxN7RLyvBTQHorsEURqnbM3VS9as0/FX5bWK62rxFvIKgLufSaxiXMRj476FgG8TD+m5NOehuzzZr9eBeF21Aq/Lkf16nZfm8lRzXqvwmvsaGITjX5Zqrqs8mu8G/gIcbWaXEsvzv+LuJxOzC5ci6u2WqTmGSLqX7XVJYv+NMjWXIs2GLll3DOWf16q8VnUNZNUciO6SFXhdinzX653EIMJ4YLS7n+juT3iUyRtLzntARbpz01ya/DFrXrp54kBV5xV3v5uYsb0QcL6732yxiuWPHntTZWJI17BOow2Pm9mBxAdxdHq9iwhKXwJW8LSRx2Bpymt9NOvmdS5/bwGiRMqhwO7unicwzxddeZXXVtW0WIa2ArGPwRtpdsEiwEXE7J093X3DwdaUV3mtodcOopZuY/PO5g1SVwEucve1BluzTl7b/fjldTb9TYnVEU8SiYUtzWxxd88zcFeprrzWx2u7Hr9F/dfFgPcTK6Ae8VhxhMXKwEvcfc3B1pRXea2b1/T7WxB7hK0N3AH8iNiDYpK75xq0q0q3Xb3a2/faWsybarhbrLC5GNjF3f+d1eeQnWFtMdNnAzO7jEj2Xd94z6McxBHuPjJjUrF0TXmtj2bdvCbtYen/Fc3si43vPWpIvULMgru4FXTlVV7rotnQ9Vgu+SapRIdHPeyXiI18zgEuHWxNeZXXmnrtBZZPbbbD315H8EiiNvCgatbJa7sfv7y+LRauZGb7uftviFUlE4gVSJBjNUQVuvJaH6/tfvwNXY/l+KOJWYkHAT9s+pGvk7GGfxWa8iqvdfLa1F67iZnJNwG3EzO1LyHi4FFZNKvSlde39tfpsNhj4EIz+62ZbZ3e+zewv+dIVsMQTlgDM4ml46OIWXSHmNkEi100MbOrzCxrHZUqNOW1Ppp18wppY0miBEKHRyH8DptVD/sE8m1gWIWuvMprXTSbdb9IbJL6VieA2AX5euCnLaApr/JaV6+NNtvoBA+zmKXzb2IWyGBr1slrux+/vM7S3JuYQQVRv3c3d7/VmmZHtYCuvNbHa7sff7PufkCPx8zCqRDJcaLU1slz+N35qSmv8lonr83tdaq7X06UxPsGMdloP3f/eUbNqnTb2qvNqnd+IFHO7HyiDNQ5Znafme3l7n/P6PMt8hQ+rwUeS+ges5i1+lNic5aNgZ+Y2cLEB/SfwdaU1/po1s1ro+NlUfuwi1l1jzuBnvS3r20FXXmV17po9qPb2aQ7zMw6PXaH3sczbNhShaa8ymvNvb6tzSat+8zsmAJeS9Gsk9d2P3557VfzXwAeq4wuT19nTtRVoSuv9fHa7sffj24n8FB6qxOY4VGD/psF42thTXmV1zp57UfzMQCPlUYN/TcGqlelrrzG6sr05a7A5sB3gfHEJuxXExvlnptVt8GQrGFtZku4+wtmthewqrsfbmajiAL6RixXuN3dB7xxWRWa8lofzbp57fM3diQ2mHNgZ49C+G/dtFpJV17ltS6ac9MtQhWaVenKq7zKa328tvvxV6U7FLy2S19AXuujOVS8FmEo3FtaUVde6+NV95b6eDWzlYEdidVfN7j7h9PrpwFHu/uTeXRhCCasLZaj7gasCWwCHO/uFza9PwaY4mkTk8HSlNf6aNbNa5+/sQCxVHsksextb+BP6W9d3Uq68iqvddGUV3mVV3mtk9d2P355lVd5rY+mvMqrvMprux9/Hbya2XBgSXd/On2/GDCd2BvqRmAy8BV3XyOP1wZDroa1R8H3a4ClgXcCG5vZLhY7agP8gKipMqia8lofzbp5BTCz5dKXewOHuPsUd98PWAC4BZhkZsu0gq68ymtdNOVVXuVVXuvktd2PX17lVV7roymv8iqv8trux18zr7sD95jZ181shaT5KnAhsZnt9sQGnIUYijOsR7n7m2a2VnppdWAv4D7gNeBT7p4pCViFprzWR7OGXkcCOwBbA2sD27v7X7LqzA9deZXXumjKq7zKq7zWyWu7H7+8yqu81kdTXuVVXuW13Y+/bl6T9ieJTRuXIepVn+Pu/0zvLeTurxX9G0NqhrWZ7Qpcb2bnEyUWdiMSfxOBdYHFiYTgoGrKa3006+Y1MRO4FRhFbCxyiJlNMLN3pb99lZkt2yK68iqvddGUV3mVV3mtk9d2P355lVd5rY+mvMqrvMprux9/bbyaWVf68m7gOGA/YHngajM7y8w2KCNZDUNshrXNSv79FngAeBH4PvEBvQCs4O7rDLamvNZHs25e+/yN3YEpwMLAxsDY9PVUd9+6lXTlVV7roimv8iqv8lonr+1+/PIqr/JaH015lVd5ldd2P/46eLW0QaOZXQNc7e6npddXBo4GlvO08WJRuub9I7Xi/4BjiY3qjgUwsz2Bk9z9eovp8K2gKa/10ayVVzNbwt1fMLO9gFXd/XAzGwXcAxgwGri9FXTlVV7roimv8iqv8lonr+1+/PIqr/JaH015lVd5ldd2P/66efVIVncDY939NDMbBoxw9/vN7IfAM1k158SQmWFts7L8BpwJPEzMXN3C3dczs+HuPn2wNeW1Ppo19DqMKC2yJjF7+3h3v7Dp/TFEgjyr39J15VVe66Ipr/Iqr/JaJ6/tfvzyKq/yWh9NeZVXeZXXdj/+unlt+v3VgK+4+25Nry0P3ACs4u4z8uj2ZcjUsHb33vS/A5sBU4GvAcekH+lpBU15rY9mDb3OBK4BlgbeCWxsZruY2SrpR34AZN7EsQpdeZXXumjKq7zKq7zWyWu7H7+8yqu81kdTXuVVXuW13Y+/bl6beARY0cyuNLONzGwRIqf187KS1TCEZlj3xcxWBU4CXgcmuPsjrahZlW67e23H4zezUe7+ppmtlV5andi48T5iQ8dPuXvmm1IVuvIqr3XRlFd5lVd5rZPXdj9+eZVXea2PprzKq7zKa7sff9289vkb7wDGAxsC7wYuB77j7lOL6DYzZGZY98Xd7wM+CywIlFLwuwrNqnTb3Wu7Hb+Z7Qpcb7M2c9yNuBFNBNYFFiduUIOuK6/yWhdNeZVXeZXXOnlt9+OXV3mV1/poyqu8yqu8tvvx181rX9z9JeB44NPAx939sDKT1TCEE9YA7v4KsAtwXStrVqXb7l7b7Pg3AVZMX3cCDlwIbEWMoq3m7n9tEV15lde6aMqrvMqrvNbJa7sfv7zKq7zWR1Ne5VVe5bXdj79uXmfD3f/r7q+5+2Nl6M1Gb2+v/umf/tX8X3d3d2d3d/cJ3d3dX2167c/d3d2fTF+PbBVdeZXXumjKq7zKq7zWyWu7H7+8yqu81kdTXuVVXuW13Y+/bl4H49+QrWEtRLtgZh3u3mtmBpwJPAw8AGzh7uuZ2XDPsftrFbryKq910ZRXeZVXea2T13Y/fnmVV3mtj6a8yqu8ymu7H3/dvA4WQ7okiBDtgLv3pv8d2AyYSuzQekz6kZ5W0ZVXea2LprzKq7zKa528tvvxy6u8ymt9NOVVXuVVXtv9+OvmdbDQDGshhhhmtipwEvA6MMHdH2lVXXmV17poVqUrr/Iqr/JaF82qdOVVXuW1Pl7b/fir0pVXeW13r+1+/FXpVuV1fqEZ1kIMMdz9PuCzwILAh1tZV17ltS6aVenKq7zKq7zWRbMqXXmVV3mtj9d2P/6qdOVVXtvda7sff1W6VXmdX2iGtRBDFDNbDJjp7lNbXVde5bUumlXpyqu8yqu81kWzKl15lVd5rY/Xdj/+qnTlVV7b3Wu7H39VulV5rRolrIUQQgghhBBCCCGEEEK0BCoJIoQQQgghhBBCCCGEEKIlUMJaCCGEEEIIIYQQQgghREughLUQQgghhBBCCCGEEEKIlkAJayGEEEIIIYQQQgghhBAtgRLWQgghhBBCCCGEEEIIIVoCJayFEEIIIYQQQgghhBBCtAT/D28fnHHSykgmAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 1800x504 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DKarHTgB-Y4N",
        "outputId": "5dcb5a58-eb96-4e7b-f495-87a85e89d446"
      },
      "source": [
        "total_cells= np.product(df_test.shape)\n",
        "total_missing=df_test.isnull().sum().sum()\n",
        "# percent of data that is missing\n",
        "percent_missing = (total_missing/total_cells) * 100\n",
        "print(percent_missing)"
      ],
      "id": "DKarHTgB-Y4N",
      "execution_count": 93,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1.3788335847159376\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 426
        },
        "id": "365181c0",
        "outputId": "38c2b7b4-e59f-48ed-cf7e-6c74a7e2da86"
      },
      "source": [
        "plt.figure(figsize = (25,7))\n",
        "plt.xticks(rotation=65)\n",
        "plt.yticks(rotation=40)\n",
        "sns.set_theme(style = \"ticks\", palette = \"pastel\")\n",
        "sns.barplot(x = df_test.columns,y = df_test.isna().sum().values)\n",
        "plt.title(\"Missing values in Test Dataset\")"
      ],
      "id": "365181c0",
      "execution_count": 94,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Missing values in Test Dataset')"
            ]
          },
          "metadata": {},
          "execution_count": 94
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1800x504 with 1 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 742
        },
        "id": "d81a4451",
        "outputId": "9ebda384-1ac7-4f94-c269-ae8fc32e20e3"
      },
      "source": [
        "plt.figure(figsize=(50,60))\n",
        "pltmv=sns.displot(data=df_train.isna().melt(value_name=\"missing\"),\n",
        "            y=\"variable\",hue=\"missing\",\n",
        "            multiple=\"fill\",height=10,aspect=1)\n",
        "plt.savefig(\"visualizing_missing_data_with_barplot.png\", dpi=100)"
      ],
      "id": "d81a4451",
      "execution_count": 95,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 3600x4320 with 0 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 787.475x720 with 1 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 775
        },
        "id": "07d3435f",
        "outputId": "7de5c3f9-9708-4dfd-c13b-056ff0cf9d3a"
      },
      "source": [
        "plt.figure(figsize=(50,60))\n",
        "pltmv=sns.displot(data=df_test.isna().melt(value_name=\"missing\"),\n",
        "            y=\"variable\",hue=\"missing\",\n",
        "            multiple=\"fill\",height=10,aspect=1)\n",
        "plt.savefig(\"visualizing_missing_data_with_barplot.png\", dpi=100)\n",
        "plt.title(\"Missing values in Test Dataset\")"
      ],
      "id": "07d3435f",
      "execution_count": 96,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Missing values in Test Dataset')"
            ]
          },
          "metadata": {},
          "execution_count": 96
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 3600x4320 with 0 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 787.475x720 with 1 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5034dbb5"
      },
      "source": [
        "Note:\n",
        "\n",
        "As we can see there are plethora of missing values in our dataset. We can use bfill, ffill, mean, mode or median for filling the missing values using fillna() method."
      ],
      "id": "5034dbb5"
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 512
        },
        "id": "xT_t7p8H-1cO",
        "outputId": "d68e172f-3f78-4ee0-f6e2-c605c0cdd563"
      },
      "source": [
        "msno.matrix(df_train, figsize=(16,8))\n",
        "plt.title(\"Missing data distribution\", size=15)"
      ],
      "id": "xT_t7p8H-1cO",
      "execution_count": 97,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Missing data distribution')"
            ]
          },
          "metadata": {},
          "execution_count": 97
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1152x576 with 2 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RzObKIYbCnAn"
      },
      "source": [
        "# *Feature Selection*\n",
        "\n",
        "---\n",
        "\n"
      ],
      "id": "RzObKIYbCnAn"
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 241
        },
        "id": "rSCUaU_DDqou",
        "outputId": "95286999-10ee-4654-8e45-09e6a8ffcad5"
      },
      "source": [
        "# concat all dataset\n",
        "df = pd.concat([df_train, df_test],axis = 0)\n",
        "\n",
        "\n",
        "#class /label encoding\n",
        "df['class']= label_encoder.fit_transform(df['class'])\n",
        "print(df['class'].unique())\n",
        "\n",
        "\n",
        "#Feature x,y\n",
        "X= df.drop('class',axis=1)\n",
        "y=df['class']\n",
        "\n",
        "df.head()"
      ],
      "id": "rSCUaU_DDqou",
      "execution_count": 98,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[0 1]\n"
          ]
        },
        {
          "output_type": "execute_result",
          "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>Attr1</th>\n",
              "      <th>Attr2</th>\n",
              "      <th>Attr3</th>\n",
              "      <th>Attr4</th>\n",
              "      <th>Attr5</th>\n",
              "      <th>Attr6</th>\n",
              "      <th>Attr7</th>\n",
              "      <th>Attr8</th>\n",
              "      <th>Attr9</th>\n",
              "      <th>Attr10</th>\n",
              "      <th>Attr11</th>\n",
              "      <th>Attr12</th>\n",
              "      <th>Attr13</th>\n",
              "      <th>Attr14</th>\n",
              "      <th>Attr15</th>\n",
              "      <th>Attr16</th>\n",
              "      <th>Attr17</th>\n",
              "      <th>Attr18</th>\n",
              "      <th>Attr19</th>\n",
              "      <th>Attr20</th>\n",
              "      <th>Attr21</th>\n",
              "      <th>Attr22</th>\n",
              "      <th>Attr23</th>\n",
              "      <th>Attr24</th>\n",
              "      <th>Attr25</th>\n",
              "      <th>Attr26</th>\n",
              "      <th>Attr27</th>\n",
              "      <th>Attr28</th>\n",
              "      <th>Attr29</th>\n",
              "      <th>Attr30</th>\n",
              "      <th>Attr31</th>\n",
              "      <th>Attr32</th>\n",
              "      <th>Attr33</th>\n",
              "      <th>Attr34</th>\n",
              "      <th>Attr35</th>\n",
              "      <th>Attr36</th>\n",
              "      <th>Attr37</th>\n",
              "      <th>Attr38</th>\n",
              "      <th>Attr39</th>\n",
              "      <th>Attr40</th>\n",
              "      <th>Attr41</th>\n",
              "      <th>Attr42</th>\n",
              "      <th>Attr43</th>\n",
              "      <th>Attr44</th>\n",
              "      <th>Attr45</th>\n",
              "      <th>Attr46</th>\n",
              "      <th>Attr47</th>\n",
              "      <th>Attr48</th>\n",
              "      <th>Attr49</th>\n",
              "      <th>Attr50</th>\n",
              "      <th>Attr51</th>\n",
              "      <th>Attr52</th>\n",
              "      <th>Attr53</th>\n",
              "      <th>Attr54</th>\n",
              "      <th>Attr55</th>\n",
              "      <th>Attr56</th>\n",
              "      <th>Attr57</th>\n",
              "      <th>Attr58</th>\n",
              "      <th>Attr59</th>\n",
              "      <th>Attr60</th>\n",
              "      <th>Attr61</th>\n",
              "      <th>Attr62</th>\n",
              "      <th>Attr63</th>\n",
              "      <th>Attr64</th>\n",
              "      <th>class</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.200550</td>\n",
              "      <td>0.37951</td>\n",
              "      <td>0.39641</td>\n",
              "      <td>2.0472</td>\n",
              "      <td>32.3510</td>\n",
              "      <td>0.38825</td>\n",
              "      <td>0.249760</td>\n",
              "      <td>1.33050</td>\n",
              "      <td>1.1389</td>\n",
              "      <td>0.50494</td>\n",
              "      <td>0.249760</td>\n",
              "      <td>0.65980</td>\n",
              "      <td>0.166600</td>\n",
              "      <td>0.249760</td>\n",
              "      <td>497.42</td>\n",
              "      <td>0.73378</td>\n",
              "      <td>2.6349</td>\n",
              "      <td>0.249760</td>\n",
              "      <td>0.149420</td>\n",
              "      <td>43.370</td>\n",
              "      <td>1.2479</td>\n",
              "      <td>0.21402</td>\n",
              "      <td>0.119980</td>\n",
              "      <td>0.47706</td>\n",
              "      <td>0.50494</td>\n",
              "      <td>0.60411</td>\n",
              "      <td>1.45820</td>\n",
              "      <td>1.7615</td>\n",
              "      <td>5.9443</td>\n",
              "      <td>0.11788</td>\n",
              "      <td>0.149420</td>\n",
              "      <td>94.14</td>\n",
              "      <td>3.8772</td>\n",
              "      <td>0.56393</td>\n",
              "      <td>0.21402</td>\n",
              "      <td>1.7410</td>\n",
              "      <td>593.2700</td>\n",
              "      <td>0.50591</td>\n",
              "      <td>0.128040</td>\n",
              "      <td>0.662950</td>\n",
              "      <td>0.051402</td>\n",
              "      <td>0.128040</td>\n",
              "      <td>114.42</td>\n",
              "      <td>71.050</td>\n",
              "      <td>1.00970</td>\n",
              "      <td>1.52250</td>\n",
              "      <td>49.394</td>\n",
              "      <td>0.185300</td>\n",
              "      <td>0.110850</td>\n",
              "      <td>2.0420</td>\n",
              "      <td>0.37854</td>\n",
              "      <td>0.25792</td>\n",
              "      <td>2.2437</td>\n",
              "      <td>2.2480</td>\n",
              "      <td>348690.0</td>\n",
              "      <td>0.121960</td>\n",
              "      <td>0.39718</td>\n",
              "      <td>0.87804</td>\n",
              "      <td>0.001924</td>\n",
              "      <td>8.4160</td>\n",
              "      <td>5.1372</td>\n",
              "      <td>82.658</td>\n",
              "      <td>4.4158</td>\n",
              "      <td>7.4277</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.209120</td>\n",
              "      <td>0.49988</td>\n",
              "      <td>0.47225</td>\n",
              "      <td>1.9447</td>\n",
              "      <td>14.7860</td>\n",
              "      <td>0.00000</td>\n",
              "      <td>0.258340</td>\n",
              "      <td>0.99601</td>\n",
              "      <td>1.6996</td>\n",
              "      <td>0.49788</td>\n",
              "      <td>0.261140</td>\n",
              "      <td>0.51680</td>\n",
              "      <td>0.158350</td>\n",
              "      <td>0.258340</td>\n",
              "      <td>677.96</td>\n",
              "      <td>0.53838</td>\n",
              "      <td>2.0005</td>\n",
              "      <td>0.258340</td>\n",
              "      <td>0.152000</td>\n",
              "      <td>87.981</td>\n",
              "      <td>1.4293</td>\n",
              "      <td>0.24806</td>\n",
              "      <td>0.123040</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.39542</td>\n",
              "      <td>0.43992</td>\n",
              "      <td>88.44400</td>\n",
              "      <td>16.9460</td>\n",
              "      <td>3.6884</td>\n",
              "      <td>0.26969</td>\n",
              "      <td>0.152000</td>\n",
              "      <td>122.17</td>\n",
              "      <td>2.9876</td>\n",
              "      <td>2.98760</td>\n",
              "      <td>0.20616</td>\n",
              "      <td>1.6996</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.49788</td>\n",
              "      <td>0.121300</td>\n",
              "      <td>0.086422</td>\n",
              "      <td>0.064371</td>\n",
              "      <td>0.145950</td>\n",
              "      <td>199.49</td>\n",
              "      <td>111.510</td>\n",
              "      <td>0.51045</td>\n",
              "      <td>1.12520</td>\n",
              "      <td>100.130</td>\n",
              "      <td>0.237270</td>\n",
              "      <td>0.139610</td>\n",
              "      <td>1.9447</td>\n",
              "      <td>0.49988</td>\n",
              "      <td>0.33472</td>\n",
              "      <td>17.8660</td>\n",
              "      <td>17.8660</td>\n",
              "      <td>2304.6</td>\n",
              "      <td>0.121300</td>\n",
              "      <td>0.42002</td>\n",
              "      <td>0.85300</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>4.1486</td>\n",
              "      <td>3.2732</td>\n",
              "      <td>107.350</td>\n",
              "      <td>3.4000</td>\n",
              "      <td>60.9870</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.248660</td>\n",
              "      <td>0.69592</td>\n",
              "      <td>0.26713</td>\n",
              "      <td>1.5548</td>\n",
              "      <td>-1.1523</td>\n",
              "      <td>0.00000</td>\n",
              "      <td>0.309060</td>\n",
              "      <td>0.43695</td>\n",
              "      <td>1.3090</td>\n",
              "      <td>0.30408</td>\n",
              "      <td>0.312580</td>\n",
              "      <td>0.64184</td>\n",
              "      <td>0.244350</td>\n",
              "      <td>0.309060</td>\n",
              "      <td>794.16</td>\n",
              "      <td>0.45961</td>\n",
              "      <td>1.4369</td>\n",
              "      <td>0.309060</td>\n",
              "      <td>0.236100</td>\n",
              "      <td>73.133</td>\n",
              "      <td>1.4283</td>\n",
              "      <td>0.30260</td>\n",
              "      <td>0.189960</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.28932</td>\n",
              "      <td>0.37282</td>\n",
              "      <td>86.01100</td>\n",
              "      <td>1.0627</td>\n",
              "      <td>4.3749</td>\n",
              "      <td>0.41929</td>\n",
              "      <td>0.238150</td>\n",
              "      <td>176.93</td>\n",
              "      <td>2.0630</td>\n",
              "      <td>1.42740</td>\n",
              "      <td>0.31565</td>\n",
              "      <td>1.3090</td>\n",
              "      <td>2.3019</td>\n",
              "      <td>0.51537</td>\n",
              "      <td>0.241140</td>\n",
              "      <td>0.322020</td>\n",
              "      <td>0.074020</td>\n",
              "      <td>0.231170</td>\n",
              "      <td>165.51</td>\n",
              "      <td>92.381</td>\n",
              "      <td>0.94807</td>\n",
              "      <td>1.01010</td>\n",
              "      <td>96.372</td>\n",
              "      <td>0.291810</td>\n",
              "      <td>0.222930</td>\n",
              "      <td>1.0758</td>\n",
              "      <td>0.48152</td>\n",
              "      <td>0.48474</td>\n",
              "      <td>1.2098</td>\n",
              "      <td>2.0504</td>\n",
              "      <td>6332.7</td>\n",
              "      <td>0.241140</td>\n",
              "      <td>0.81774</td>\n",
              "      <td>0.76599</td>\n",
              "      <td>0.694840</td>\n",
              "      <td>4.9909</td>\n",
              "      <td>3.9510</td>\n",
              "      <td>134.270</td>\n",
              "      <td>2.7185</td>\n",
              "      <td>5.2078</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.081483</td>\n",
              "      <td>0.30734</td>\n",
              "      <td>0.45879</td>\n",
              "      <td>2.4928</td>\n",
              "      <td>51.9520</td>\n",
              "      <td>0.14988</td>\n",
              "      <td>0.092704</td>\n",
              "      <td>1.86610</td>\n",
              "      <td>1.0571</td>\n",
              "      <td>0.57353</td>\n",
              "      <td>0.092704</td>\n",
              "      <td>0.30163</td>\n",
              "      <td>0.094257</td>\n",
              "      <td>0.092704</td>\n",
              "      <td>917.01</td>\n",
              "      <td>0.39803</td>\n",
              "      <td>3.2537</td>\n",
              "      <td>0.092704</td>\n",
              "      <td>0.071428</td>\n",
              "      <td>79.788</td>\n",
              "      <td>1.5069</td>\n",
              "      <td>0.11550</td>\n",
              "      <td>0.062782</td>\n",
              "      <td>0.17193</td>\n",
              "      <td>0.57353</td>\n",
              "      <td>0.36152</td>\n",
              "      <td>0.94076</td>\n",
              "      <td>1.9618</td>\n",
              "      <td>4.6511</td>\n",
              "      <td>0.14343</td>\n",
              "      <td>0.071428</td>\n",
              "      <td>91.37</td>\n",
              "      <td>3.9948</td>\n",
              "      <td>0.37581</td>\n",
              "      <td>0.11550</td>\n",
              "      <td>1.3562</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.57353</td>\n",
              "      <td>0.088995</td>\n",
              "      <td>0.401390</td>\n",
              "      <td>0.069622</td>\n",
              "      <td>0.088995</td>\n",
              "      <td>180.77</td>\n",
              "      <td>100.980</td>\n",
              "      <td>0.28720</td>\n",
              "      <td>1.56960</td>\n",
              "      <td>84.344</td>\n",
              "      <td>0.085874</td>\n",
              "      <td>0.066165</td>\n",
              "      <td>2.4928</td>\n",
              "      <td>0.30734</td>\n",
              "      <td>0.25033</td>\n",
              "      <td>2.4524</td>\n",
              "      <td>2.4524</td>\n",
              "      <td>20545.0</td>\n",
              "      <td>0.054015</td>\n",
              "      <td>0.14207</td>\n",
              "      <td>0.94598</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>4.5746</td>\n",
              "      <td>3.6147</td>\n",
              "      <td>86.435</td>\n",
              "      <td>4.2228</td>\n",
              "      <td>5.5497</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.187320</td>\n",
              "      <td>0.61323</td>\n",
              "      <td>0.22960</td>\n",
              "      <td>1.4063</td>\n",
              "      <td>-7.3128</td>\n",
              "      <td>0.18732</td>\n",
              "      <td>0.187320</td>\n",
              "      <td>0.63070</td>\n",
              "      <td>1.1559</td>\n",
              "      <td>0.38677</td>\n",
              "      <td>0.187320</td>\n",
              "      <td>0.33147</td>\n",
              "      <td>0.121820</td>\n",
              "      <td>0.187320</td>\n",
              "      <td>1133.20</td>\n",
              "      <td>0.32211</td>\n",
              "      <td>1.6307</td>\n",
              "      <td>0.187320</td>\n",
              "      <td>0.115530</td>\n",
              "      <td>57.045</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.19832</td>\n",
              "      <td>0.115530</td>\n",
              "      <td>0.18732</td>\n",
              "      <td>0.38677</td>\n",
              "      <td>0.32211</td>\n",
              "      <td>1.41380</td>\n",
              "      <td>1.1184</td>\n",
              "      <td>4.1424</td>\n",
              "      <td>0.27884</td>\n",
              "      <td>0.115530</td>\n",
              "      <td>147.04</td>\n",
              "      <td>2.4823</td>\n",
              "      <td>0.32340</td>\n",
              "      <td>0.19832</td>\n",
              "      <td>1.6278</td>\n",
              "      <td>11.2470</td>\n",
              "      <td>0.43489</td>\n",
              "      <td>0.122310</td>\n",
              "      <td>0.293040</td>\n",
              "      <td>0.096680</td>\n",
              "      <td>0.122310</td>\n",
              "      <td>141.62</td>\n",
              "      <td>84.574</td>\n",
              "      <td>0.73919</td>\n",
              "      <td>0.95787</td>\n",
              "      <td>65.936</td>\n",
              "      <td>0.188110</td>\n",
              "      <td>0.116010</td>\n",
              "      <td>1.2959</td>\n",
              "      <td>0.56511</td>\n",
              "      <td>0.40285</td>\n",
              "      <td>1.8839</td>\n",
              "      <td>2.1184</td>\n",
              "      <td>3186.6</td>\n",
              "      <td>0.134850</td>\n",
              "      <td>0.48431</td>\n",
              "      <td>0.86515</td>\n",
              "      <td>0.124440</td>\n",
              "      <td>6.3985</td>\n",
              "      <td>4.3158</td>\n",
              "      <td>127.210</td>\n",
              "      <td>2.8692</td>\n",
              "      <td>7.8980</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      Attr1    Attr2    Attr3   Attr4  ...   Attr62  Attr63   Attr64  class\n",
              "0  0.200550  0.37951  0.39641  2.0472  ...   82.658  4.4158   7.4277      0\n",
              "1  0.209120  0.49988  0.47225  1.9447  ...  107.350  3.4000  60.9870      0\n",
              "2  0.248660  0.69592  0.26713  1.5548  ...  134.270  2.7185   5.2078      0\n",
              "3  0.081483  0.30734  0.45879  2.4928  ...   86.435  4.2228   5.5497      0\n",
              "4  0.187320  0.61323  0.22960  1.4063  ...  127.210  2.8692   7.8980      0\n",
              "\n",
              "[5 rows x 65 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 98
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 302
        },
        "id": "zNWImq2bIL08",
        "outputId": "94b802f2-ab84-4fec-b2e2-751149f23c9e"
      },
      "source": [
        "sns.countplot(df['class'])"
      ],
      "id": "zNWImq2bIL08",
      "execution_count": 99,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f0780a81710>"
            ]
          },
          "metadata": {},
          "execution_count": 99
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PBWDOc9mJ04j"
      },
      "source": [
        "# Data Imputation\n",
        "\n",
        "---\n",
        "\n"
      ],
      "id": "PBWDOc9mJ04j"
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_jlgr_TmCiHn"
      },
      "source": [
        "from sklearn.impute import SimpleImputer\n",
        "\n",
        "imputer = SimpleImputer(strategy=\"mean\")\n",
        "X_imp = imputer.fit_transform(X)"
      ],
      "id": "_jlgr_TmCiHn",
      "execution_count": 100,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JFtbA5rABVfh",
        "outputId": "eacf468c-628f-4d7b-ae5e-df01b55145b6"
      },
      "source": [
        "from sklearn.feature_selection import RFE\n",
        "import lightgbm as lgb\n",
        "#Feature selection \n",
        "#Recursive feature elimination\n",
        "\n",
        "classifier = lgb.LGBMClassifier()\n",
        "#Select best feature \n",
        "rfe = RFE(classifier,20)\n",
        "rfe = rfe.fit(X_imp, y )\n",
        "\n",
        "#Summarize the selection of the attributes\n",
        "print(rfe.support_)\n",
        "print(rfe.ranking_)\n",
        "pd.DataFrame(X_imp,columns=data_keys[:len(data_keys)-1]).columns[rfe.support_]"
      ],
      "id": "JFtbA5rABVfh",
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[False False False  True  True  True False False False False False False\n",
            "  True False False False False False False  True False  True False  True\n",
            "  True False  True False  True  True False False False  True False  True\n",
            " False False  True  True False False False  True False  True False False\n",
            " False False False False False False  True  True False  True False False\n",
            " False False False False]\n",
            "[36 42 13  1  1  1 40 26  6 41 25 33  1 45 35 30 43 44  5  1  2  1 39  1\n",
            "  1 14  1 29  1  1 28 15  7  1  8  1 12  3  1  1 10 21 23  1 24  1  4 18\n",
            " 27 20 37 31 32 16  1  1 22  1 34 17 11 19 38  9]\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['Attr4', 'Attr5', 'Attr6', 'Attr13', 'Attr20', 'Attr22', 'Attr24',\n",
              "       'Attr25', 'Attr27', 'Attr29', 'Attr30', 'Attr34', 'Attr36', 'Attr39',\n",
              "       'Attr40', 'Attr44', 'Attr46', 'Attr55', 'Attr56', 'Attr58'],\n",
              "      dtype='object')"
            ]
          },
          "metadata": {},
          "execution_count": 27
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 411
        },
        "id": "4bi9Iu86Jau8",
        "outputId": "a583b552-5129-41f4-c9c7-daaf22857333"
      },
      "source": [
        "X_imp= pd.DataFrame(X_imp,columns=data_keys[:len(data_keys)-1])\n",
        "X_fs= X_imp[X_imp.columns[rfe.support_]]\n",
        "X_fs.head()"
      ],
      "id": "4bi9Iu86Jau8",
      "execution_count": 101,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>Attr4</th>\n",
              "      <th>Attr5</th>\n",
              "      <th>Attr6</th>\n",
              "      <th>Attr13</th>\n",
              "      <th>Attr20</th>\n",
              "      <th>Attr22</th>\n",
              "      <th>Attr24</th>\n",
              "      <th>Attr25</th>\n",
              "      <th>Attr27</th>\n",
              "      <th>Attr29</th>\n",
              "      <th>Attr30</th>\n",
              "      <th>Attr34</th>\n",
              "      <th>Attr36</th>\n",
              "      <th>Attr39</th>\n",
              "      <th>Attr40</th>\n",
              "      <th>Attr44</th>\n",
              "      <th>Attr46</th>\n",
              "      <th>Attr55</th>\n",
              "      <th>Attr56</th>\n",
              "      <th>Attr58</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2.0472</td>\n",
              "      <td>32.3510</td>\n",
              "      <td>0.38825</td>\n",
              "      <td>0.166600</td>\n",
              "      <td>43.370</td>\n",
              "      <td>0.21402</td>\n",
              "      <td>0.477060</td>\n",
              "      <td>0.50494</td>\n",
              "      <td>1.45820</td>\n",
              "      <td>5.9443</td>\n",
              "      <td>0.11788</td>\n",
              "      <td>0.56393</td>\n",
              "      <td>1.7410</td>\n",
              "      <td>0.128040</td>\n",
              "      <td>0.662950</td>\n",
              "      <td>71.050</td>\n",
              "      <td>1.52250</td>\n",
              "      <td>348690.0</td>\n",
              "      <td>0.121960</td>\n",
              "      <td>0.87804</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1.9447</td>\n",
              "      <td>14.7860</td>\n",
              "      <td>0.00000</td>\n",
              "      <td>0.158350</td>\n",
              "      <td>87.981</td>\n",
              "      <td>0.24806</td>\n",
              "      <td>0.270012</td>\n",
              "      <td>0.39542</td>\n",
              "      <td>88.44400</td>\n",
              "      <td>3.6884</td>\n",
              "      <td>0.26969</td>\n",
              "      <td>2.98760</td>\n",
              "      <td>1.6996</td>\n",
              "      <td>0.121300</td>\n",
              "      <td>0.086422</td>\n",
              "      <td>111.510</td>\n",
              "      <td>1.12520</td>\n",
              "      <td>2304.6</td>\n",
              "      <td>0.121300</td>\n",
              "      <td>0.85300</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1.5548</td>\n",
              "      <td>-1.1523</td>\n",
              "      <td>0.00000</td>\n",
              "      <td>0.244350</td>\n",
              "      <td>73.133</td>\n",
              "      <td>0.30260</td>\n",
              "      <td>0.270012</td>\n",
              "      <td>0.28932</td>\n",
              "      <td>86.01100</td>\n",
              "      <td>4.3749</td>\n",
              "      <td>0.41929</td>\n",
              "      <td>1.42740</td>\n",
              "      <td>1.3090</td>\n",
              "      <td>0.241140</td>\n",
              "      <td>0.322020</td>\n",
              "      <td>92.381</td>\n",
              "      <td>1.01010</td>\n",
              "      <td>6332.7</td>\n",
              "      <td>0.241140</td>\n",
              "      <td>0.76599</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>2.4928</td>\n",
              "      <td>51.9520</td>\n",
              "      <td>0.14988</td>\n",
              "      <td>0.094257</td>\n",
              "      <td>79.788</td>\n",
              "      <td>0.11550</td>\n",
              "      <td>0.171930</td>\n",
              "      <td>0.57353</td>\n",
              "      <td>0.94076</td>\n",
              "      <td>4.6511</td>\n",
              "      <td>0.14343</td>\n",
              "      <td>0.37581</td>\n",
              "      <td>1.3562</td>\n",
              "      <td>0.088995</td>\n",
              "      <td>0.401390</td>\n",
              "      <td>100.980</td>\n",
              "      <td>1.56960</td>\n",
              "      <td>20545.0</td>\n",
              "      <td>0.054015</td>\n",
              "      <td>0.94598</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1.4063</td>\n",
              "      <td>-7.3128</td>\n",
              "      <td>0.18732</td>\n",
              "      <td>0.121820</td>\n",
              "      <td>57.045</td>\n",
              "      <td>0.19832</td>\n",
              "      <td>0.187320</td>\n",
              "      <td>0.38677</td>\n",
              "      <td>1.41380</td>\n",
              "      <td>4.1424</td>\n",
              "      <td>0.27884</td>\n",
              "      <td>0.32340</td>\n",
              "      <td>1.6278</td>\n",
              "      <td>0.122310</td>\n",
              "      <td>0.293040</td>\n",
              "      <td>84.574</td>\n",
              "      <td>0.95787</td>\n",
              "      <td>3186.6</td>\n",
              "      <td>0.134850</td>\n",
              "      <td>0.86515</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "    Attr4    Attr5    Attr6    Attr13  ...   Attr46    Attr55    Attr56   Attr58\n",
              "0  2.0472  32.3510  0.38825  0.166600  ...  1.52250  348690.0  0.121960  0.87804\n",
              "1  1.9447  14.7860  0.00000  0.158350  ...  1.12520    2304.6  0.121300  0.85300\n",
              "2  1.5548  -1.1523  0.00000  0.244350  ...  1.01010    6332.7  0.241140  0.76599\n",
              "3  2.4928  51.9520  0.14988  0.094257  ...  1.56960   20545.0  0.054015  0.94598\n",
              "4  1.4063  -7.3128  0.18732  0.121820  ...  0.95787    3186.6  0.134850  0.86515\n",
              "\n",
              "[5 rows x 20 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 28
        },
        {
          "output_type": "execute_result",
          "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>Attr4</th>\n",
              "      <th>Attr5</th>\n",
              "      <th>Attr6</th>\n",
              "      <th>Attr13</th>\n",
              "      <th>Attr20</th>\n",
              "      <th>Attr22</th>\n",
              "      <th>Attr24</th>\n",
              "      <th>Attr25</th>\n",
              "      <th>Attr27</th>\n",
              "      <th>Attr29</th>\n",
              "      <th>Attr30</th>\n",
              "      <th>Attr34</th>\n",
              "      <th>Attr36</th>\n",
              "      <th>Attr39</th>\n",
              "      <th>Attr40</th>\n",
              "      <th>Attr44</th>\n",
              "      <th>Attr46</th>\n",
              "      <th>Attr55</th>\n",
              "      <th>Attr56</th>\n",
              "      <th>Attr58</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2.0472</td>\n",
              "      <td>32.3510</td>\n",
              "      <td>0.38825</td>\n",
              "      <td>0.166600</td>\n",
              "      <td>43.370</td>\n",
              "      <td>0.21402</td>\n",
              "      <td>0.477060</td>\n",
              "      <td>0.50494</td>\n",
              "      <td>1.45820</td>\n",
              "      <td>5.9443</td>\n",
              "      <td>0.11788</td>\n",
              "      <td>0.56393</td>\n",
              "      <td>1.7410</td>\n",
              "      <td>0.128040</td>\n",
              "      <td>0.662950</td>\n",
              "      <td>71.050</td>\n",
              "      <td>1.52250</td>\n",
              "      <td>348690.0</td>\n",
              "      <td>0.121960</td>\n",
              "      <td>0.87804</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1.9447</td>\n",
              "      <td>14.7860</td>\n",
              "      <td>0.00000</td>\n",
              "      <td>0.158350</td>\n",
              "      <td>87.981</td>\n",
              "      <td>0.24806</td>\n",
              "      <td>0.270012</td>\n",
              "      <td>0.39542</td>\n",
              "      <td>88.44400</td>\n",
              "      <td>3.6884</td>\n",
              "      <td>0.26969</td>\n",
              "      <td>2.98760</td>\n",
              "      <td>1.6996</td>\n",
              "      <td>0.121300</td>\n",
              "      <td>0.086422</td>\n",
              "      <td>111.510</td>\n",
              "      <td>1.12520</td>\n",
              "      <td>2304.6</td>\n",
              "      <td>0.121300</td>\n",
              "      <td>0.85300</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1.5548</td>\n",
              "      <td>-1.1523</td>\n",
              "      <td>0.00000</td>\n",
              "      <td>0.244350</td>\n",
              "      <td>73.133</td>\n",
              "      <td>0.30260</td>\n",
              "      <td>0.270012</td>\n",
              "      <td>0.28932</td>\n",
              "      <td>86.01100</td>\n",
              "      <td>4.3749</td>\n",
              "      <td>0.41929</td>\n",
              "      <td>1.42740</td>\n",
              "      <td>1.3090</td>\n",
              "      <td>0.241140</td>\n",
              "      <td>0.322020</td>\n",
              "      <td>92.381</td>\n",
              "      <td>1.01010</td>\n",
              "      <td>6332.7</td>\n",
              "      <td>0.241140</td>\n",
              "      <td>0.76599</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>2.4928</td>\n",
              "      <td>51.9520</td>\n",
              "      <td>0.14988</td>\n",
              "      <td>0.094257</td>\n",
              "      <td>79.788</td>\n",
              "      <td>0.11550</td>\n",
              "      <td>0.171930</td>\n",
              "      <td>0.57353</td>\n",
              "      <td>0.94076</td>\n",
              "      <td>4.6511</td>\n",
              "      <td>0.14343</td>\n",
              "      <td>0.37581</td>\n",
              "      <td>1.3562</td>\n",
              "      <td>0.088995</td>\n",
              "      <td>0.401390</td>\n",
              "      <td>100.980</td>\n",
              "      <td>1.56960</td>\n",
              "      <td>20545.0</td>\n",
              "      <td>0.054015</td>\n",
              "      <td>0.94598</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1.4063</td>\n",
              "      <td>-7.3128</td>\n",
              "      <td>0.18732</td>\n",
              "      <td>0.121820</td>\n",
              "      <td>57.045</td>\n",
              "      <td>0.19832</td>\n",
              "      <td>0.187320</td>\n",
              "      <td>0.38677</td>\n",
              "      <td>1.41380</td>\n",
              "      <td>4.1424</td>\n",
              "      <td>0.27884</td>\n",
              "      <td>0.32340</td>\n",
              "      <td>1.6278</td>\n",
              "      <td>0.122310</td>\n",
              "      <td>0.293040</td>\n",
              "      <td>84.574</td>\n",
              "      <td>0.95787</td>\n",
              "      <td>3186.6</td>\n",
              "      <td>0.134850</td>\n",
              "      <td>0.86515</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "    Attr4    Attr5    Attr6    Attr13  ...   Attr46    Attr55    Attr56   Attr58\n",
              "0  2.0472  32.3510  0.38825  0.166600  ...  1.52250  348690.0  0.121960  0.87804\n",
              "1  1.9447  14.7860  0.00000  0.158350  ...  1.12520    2304.6  0.121300  0.85300\n",
              "2  1.5548  -1.1523  0.00000  0.244350  ...  1.01010    6332.7  0.241140  0.76599\n",
              "3  2.4928  51.9520  0.14988  0.094257  ...  1.56960   20545.0  0.054015  0.94598\n",
              "4  1.4063  -7.3128  0.18732  0.121820  ...  0.95787    3186.6  0.134850  0.86515\n",
              "\n",
              "[5 rows x 20 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 101
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "BfdoRYeKBpup",
        "outputId": "f3df6f51-9246-4c03-cd83-181474b9fa1d"
      },
      "source": [
        "# New Correlation Matrix\n",
        "sns.set(style=\"white\")\n",
        "\n",
        "# Compute the correlation matrix\n",
        "corr = X_fs.corr()\n",
        "\n",
        "# Generate a mask for the upper triangle\n",
        "mask = np.zeros_like(corr, dtype=np.bool)\n",
        "mask[np.triu_indices_from(mask)] = True\n",
        "\n",
        "# Set up the matplotlib figure\n",
        "f, ax = plt.subplots(figsize=(18, 15))\n",
        "\n",
        "# Generate a custom diverging colormap\n",
        "cmap = sns.diverging_palette(220, 10, as_cmap=True)\n",
        "\n",
        "# Draw the heatmap with the mask and correct aspect ratio\n",
        "sns.heatmap(corr, mask=mask, cmap=cmap, vmax=.3, center=0,\n",
        "            square=True, linewidths=.5, cbar_kws={\"shrink\": .5})  "
      ],
      "id": "BfdoRYeKBpup",
      "execution_count": 102,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f0783400a50>"
            ]
          },
          "metadata": {},
          "execution_count": 29
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1296x1080 with 2 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f0780c78790>"
            ]
          },
          "metadata": {},
          "execution_count": 102
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1296x1080 with 2 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "a38de37a"
      },
      "source": [
        "## Dealing with Imabalance Data \n",
        "\n",
        "---\n",
        "\n"
      ],
      "id": "a38de37a"
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 621
        },
        "id": "b77379ae",
        "outputId": "5fe51fe2-4026-4fd4-cb3c-f1c7aa33ac8b"
      },
      "source": [
        "plt.figure(figsize = (7,5))\n",
        "labels = [\"Not Bankrupt\", \"Bankrupt\"]\n",
        "plt.pie(df_train['class'].value_counts().values,labels = labels, autopct='%.1f%%')\n",
        "plt.title(\"Distribution of Bankrupt and Not-Bankrupt records in Test data\")\n",
        "plt.show()"
      ],
      "id": "b77379ae",
      "execution_count": 103,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 504x360 with 1 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 504x360 with 1 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 621
        },
        "id": "00ea3828",
        "outputId": "1c54a038-a5f8-4557-ce5d-546f33fc2055"
      },
      "source": [
        "plt.figure(figsize = (7,5))\n",
        "labels = [\"Not Bankrupt\", \"Bankrupt\"]\n",
        "plt.pie(df_test['class'].value_counts().values,labels = labels, autopct='%.1f%%')\n",
        "plt.title(\"Distribution of Bankrupt and Not-Bankrupt records in Test data\")\n",
        "plt.show()"
      ],
      "id": "00ea3828",
      "execution_count": 104,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 504x360 with 1 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 504x360 with 1 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "a194c95a"
      },
      "source": [
        "Note:\n",
        "    \n",
        "It can be clearly seen from the pie chart that the dataset is imbalanced. So we will now apply SMOTE for oversampling of data values."
      ],
      "id": "a194c95a"
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "68396f62"
      },
      "source": [
        "oversample = SMOTE()"
      ],
      "id": "68396f62",
      "execution_count": 105,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "699a5eaa"
      },
      "source": [
        "df_train= pd.DataFrame(imputer.fit_transform(df_train),columns=data_keys)\n",
        "df_test= pd.DataFrame(imputer.fit_transform(df_test),columns=data_keys)\n",
        "\n",
        "#copy of data a\n",
        "xtrain=df_train.drop('class',axis=1).copy()\n",
        "xtest=df_test.drop('class',axis=1).copy()\n",
        "y1=df_train['class']; y2=df_test['class']\n",
        "\n",
        "#oversampling \n",
        "X_train1, y_train1 = oversample.fit_resample(xtrain,y1)\n",
        "X_test1, y_test1 = oversample.fit_resample(xtest,y2)"
      ],
      "id": "699a5eaa",
      "execution_count": 106,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 621
        },
        "id": "86ffda52",
        "outputId": "fb765bbb-0f08-47ec-8dbd-e1b55870e745"
      },
      "source": [
        "plt.figure(figsize = (7,5))\n",
        "labels = [\"Not Bankrupt\", \"Bankrupt\"]\n",
        "\n",
        "plt.pie(pd.DataFrame(y_train1).value_counts(),labels = labels, autopct='%.1f%%')\n",
        "plt.title(\"Distribution of Bankrupt and Not-Bankrupt records in Training data\")\n",
        "plt.show()"
      ],
      "id": "86ffda52",
      "execution_count": 107,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 504x360 with 1 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 504x360 with 1 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0d5ef811"
      },
      "source": [
        "Note:\n",
        "\n",
        "The dataset is balanced now at the ratio of 50: 50"
      ],
      "id": "0d5ef811"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "w4ENAyP3MwpW"
      },
      "source": [
        ">> Final Data :\n",
        "\n",
        "\n",
        "1.   X_fs ->> x features selected with 24 attributes\n",
        "2.   y ->> y or label \n",
        "\n",
        "---\n",
        "\n",
        "\n",
        "\n"
      ],
      "id": "w4ENAyP3MwpW"
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 411
        },
        "id": "Y4LzGkWnMvwU",
        "outputId": "fe494503-ba59-470f-9c44-2a67e3a10e30"
      },
      "source": [
        "X_fs.head()"
      ],
      "id": "Y4LzGkWnMvwU",
      "execution_count": 108,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>Attr4</th>\n",
              "      <th>Attr5</th>\n",
              "      <th>Attr6</th>\n",
              "      <th>Attr13</th>\n",
              "      <th>Attr20</th>\n",
              "      <th>Attr22</th>\n",
              "      <th>Attr24</th>\n",
              "      <th>Attr25</th>\n",
              "      <th>Attr27</th>\n",
              "      <th>Attr29</th>\n",
              "      <th>Attr30</th>\n",
              "      <th>Attr34</th>\n",
              "      <th>Attr36</th>\n",
              "      <th>Attr39</th>\n",
              "      <th>Attr40</th>\n",
              "      <th>Attr44</th>\n",
              "      <th>Attr46</th>\n",
              "      <th>Attr55</th>\n",
              "      <th>Attr56</th>\n",
              "      <th>Attr58</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2.0472</td>\n",
              "      <td>32.3510</td>\n",
              "      <td>0.38825</td>\n",
              "      <td>0.166600</td>\n",
              "      <td>43.370</td>\n",
              "      <td>0.21402</td>\n",
              "      <td>0.477060</td>\n",
              "      <td>0.50494</td>\n",
              "      <td>1.45820</td>\n",
              "      <td>5.9443</td>\n",
              "      <td>0.11788</td>\n",
              "      <td>0.56393</td>\n",
              "      <td>1.7410</td>\n",
              "      <td>0.128040</td>\n",
              "      <td>0.662950</td>\n",
              "      <td>71.050</td>\n",
              "      <td>1.52250</td>\n",
              "      <td>348690.0</td>\n",
              "      <td>0.121960</td>\n",
              "      <td>0.87804</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1.9447</td>\n",
              "      <td>14.7860</td>\n",
              "      <td>0.00000</td>\n",
              "      <td>0.158350</td>\n",
              "      <td>87.981</td>\n",
              "      <td>0.24806</td>\n",
              "      <td>0.270012</td>\n",
              "      <td>0.39542</td>\n",
              "      <td>88.44400</td>\n",
              "      <td>3.6884</td>\n",
              "      <td>0.26969</td>\n",
              "      <td>2.98760</td>\n",
              "      <td>1.6996</td>\n",
              "      <td>0.121300</td>\n",
              "      <td>0.086422</td>\n",
              "      <td>111.510</td>\n",
              "      <td>1.12520</td>\n",
              "      <td>2304.6</td>\n",
              "      <td>0.121300</td>\n",
              "      <td>0.85300</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1.5548</td>\n",
              "      <td>-1.1523</td>\n",
              "      <td>0.00000</td>\n",
              "      <td>0.244350</td>\n",
              "      <td>73.133</td>\n",
              "      <td>0.30260</td>\n",
              "      <td>0.270012</td>\n",
              "      <td>0.28932</td>\n",
              "      <td>86.01100</td>\n",
              "      <td>4.3749</td>\n",
              "      <td>0.41929</td>\n",
              "      <td>1.42740</td>\n",
              "      <td>1.3090</td>\n",
              "      <td>0.241140</td>\n",
              "      <td>0.322020</td>\n",
              "      <td>92.381</td>\n",
              "      <td>1.01010</td>\n",
              "      <td>6332.7</td>\n",
              "      <td>0.241140</td>\n",
              "      <td>0.76599</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>2.4928</td>\n",
              "      <td>51.9520</td>\n",
              "      <td>0.14988</td>\n",
              "      <td>0.094257</td>\n",
              "      <td>79.788</td>\n",
              "      <td>0.11550</td>\n",
              "      <td>0.171930</td>\n",
              "      <td>0.57353</td>\n",
              "      <td>0.94076</td>\n",
              "      <td>4.6511</td>\n",
              "      <td>0.14343</td>\n",
              "      <td>0.37581</td>\n",
              "      <td>1.3562</td>\n",
              "      <td>0.088995</td>\n",
              "      <td>0.401390</td>\n",
              "      <td>100.980</td>\n",
              "      <td>1.56960</td>\n",
              "      <td>20545.0</td>\n",
              "      <td>0.054015</td>\n",
              "      <td>0.94598</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1.4063</td>\n",
              "      <td>-7.3128</td>\n",
              "      <td>0.18732</td>\n",
              "      <td>0.121820</td>\n",
              "      <td>57.045</td>\n",
              "      <td>0.19832</td>\n",
              "      <td>0.187320</td>\n",
              "      <td>0.38677</td>\n",
              "      <td>1.41380</td>\n",
              "      <td>4.1424</td>\n",
              "      <td>0.27884</td>\n",
              "      <td>0.32340</td>\n",
              "      <td>1.6278</td>\n",
              "      <td>0.122310</td>\n",
              "      <td>0.293040</td>\n",
              "      <td>84.574</td>\n",
              "      <td>0.95787</td>\n",
              "      <td>3186.6</td>\n",
              "      <td>0.134850</td>\n",
              "      <td>0.86515</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "    Attr4    Attr5    Attr6    Attr13  ...   Attr46    Attr55    Attr56   Attr58\n",
              "0  2.0472  32.3510  0.38825  0.166600  ...  1.52250  348690.0  0.121960  0.87804\n",
              "1  1.9447  14.7860  0.00000  0.158350  ...  1.12520    2304.6  0.121300  0.85300\n",
              "2  1.5548  -1.1523  0.00000  0.244350  ...  1.01010    6332.7  0.241140  0.76599\n",
              "3  2.4928  51.9520  0.14988  0.094257  ...  1.56960   20545.0  0.054015  0.94598\n",
              "4  1.4063  -7.3128  0.18732  0.121820  ...  0.95787    3186.6  0.134850  0.86515\n",
              "\n",
              "[5 rows x 20 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 35
        },
        {
          "output_type": "execute_result",
          "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>Attr4</th>\n",
              "      <th>Attr5</th>\n",
              "      <th>Attr6</th>\n",
              "      <th>Attr13</th>\n",
              "      <th>Attr20</th>\n",
              "      <th>Attr22</th>\n",
              "      <th>Attr24</th>\n",
              "      <th>Attr25</th>\n",
              "      <th>Attr27</th>\n",
              "      <th>Attr29</th>\n",
              "      <th>Attr30</th>\n",
              "      <th>Attr34</th>\n",
              "      <th>Attr36</th>\n",
              "      <th>Attr39</th>\n",
              "      <th>Attr40</th>\n",
              "      <th>Attr44</th>\n",
              "      <th>Attr46</th>\n",
              "      <th>Attr55</th>\n",
              "      <th>Attr56</th>\n",
              "      <th>Attr58</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2.0472</td>\n",
              "      <td>32.3510</td>\n",
              "      <td>0.38825</td>\n",
              "      <td>0.166600</td>\n",
              "      <td>43.370</td>\n",
              "      <td>0.21402</td>\n",
              "      <td>0.477060</td>\n",
              "      <td>0.50494</td>\n",
              "      <td>1.45820</td>\n",
              "      <td>5.9443</td>\n",
              "      <td>0.11788</td>\n",
              "      <td>0.56393</td>\n",
              "      <td>1.7410</td>\n",
              "      <td>0.128040</td>\n",
              "      <td>0.662950</td>\n",
              "      <td>71.050</td>\n",
              "      <td>1.52250</td>\n",
              "      <td>348690.0</td>\n",
              "      <td>0.121960</td>\n",
              "      <td>0.87804</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1.9447</td>\n",
              "      <td>14.7860</td>\n",
              "      <td>0.00000</td>\n",
              "      <td>0.158350</td>\n",
              "      <td>87.981</td>\n",
              "      <td>0.24806</td>\n",
              "      <td>0.270012</td>\n",
              "      <td>0.39542</td>\n",
              "      <td>88.44400</td>\n",
              "      <td>3.6884</td>\n",
              "      <td>0.26969</td>\n",
              "      <td>2.98760</td>\n",
              "      <td>1.6996</td>\n",
              "      <td>0.121300</td>\n",
              "      <td>0.086422</td>\n",
              "      <td>111.510</td>\n",
              "      <td>1.12520</td>\n",
              "      <td>2304.6</td>\n",
              "      <td>0.121300</td>\n",
              "      <td>0.85300</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1.5548</td>\n",
              "      <td>-1.1523</td>\n",
              "      <td>0.00000</td>\n",
              "      <td>0.244350</td>\n",
              "      <td>73.133</td>\n",
              "      <td>0.30260</td>\n",
              "      <td>0.270012</td>\n",
              "      <td>0.28932</td>\n",
              "      <td>86.01100</td>\n",
              "      <td>4.3749</td>\n",
              "      <td>0.41929</td>\n",
              "      <td>1.42740</td>\n",
              "      <td>1.3090</td>\n",
              "      <td>0.241140</td>\n",
              "      <td>0.322020</td>\n",
              "      <td>92.381</td>\n",
              "      <td>1.01010</td>\n",
              "      <td>6332.7</td>\n",
              "      <td>0.241140</td>\n",
              "      <td>0.76599</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>2.4928</td>\n",
              "      <td>51.9520</td>\n",
              "      <td>0.14988</td>\n",
              "      <td>0.094257</td>\n",
              "      <td>79.788</td>\n",
              "      <td>0.11550</td>\n",
              "      <td>0.171930</td>\n",
              "      <td>0.57353</td>\n",
              "      <td>0.94076</td>\n",
              "      <td>4.6511</td>\n",
              "      <td>0.14343</td>\n",
              "      <td>0.37581</td>\n",
              "      <td>1.3562</td>\n",
              "      <td>0.088995</td>\n",
              "      <td>0.401390</td>\n",
              "      <td>100.980</td>\n",
              "      <td>1.56960</td>\n",
              "      <td>20545.0</td>\n",
              "      <td>0.054015</td>\n",
              "      <td>0.94598</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1.4063</td>\n",
              "      <td>-7.3128</td>\n",
              "      <td>0.18732</td>\n",
              "      <td>0.121820</td>\n",
              "      <td>57.045</td>\n",
              "      <td>0.19832</td>\n",
              "      <td>0.187320</td>\n",
              "      <td>0.38677</td>\n",
              "      <td>1.41380</td>\n",
              "      <td>4.1424</td>\n",
              "      <td>0.27884</td>\n",
              "      <td>0.32340</td>\n",
              "      <td>1.6278</td>\n",
              "      <td>0.122310</td>\n",
              "      <td>0.293040</td>\n",
              "      <td>84.574</td>\n",
              "      <td>0.95787</td>\n",
              "      <td>3186.6</td>\n",
              "      <td>0.134850</td>\n",
              "      <td>0.86515</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "    Attr4    Attr5    Attr6    Attr13  ...   Attr46    Attr55    Attr56   Attr58\n",
              "0  2.0472  32.3510  0.38825  0.166600  ...  1.52250  348690.0  0.121960  0.87804\n",
              "1  1.9447  14.7860  0.00000  0.158350  ...  1.12520    2304.6  0.121300  0.85300\n",
              "2  1.5548  -1.1523  0.00000  0.244350  ...  1.01010    6332.7  0.241140  0.76599\n",
              "3  2.4928  51.9520  0.14988  0.094257  ...  1.56960   20545.0  0.054015  0.94598\n",
              "4  1.4063  -7.3128  0.18732  0.121820  ...  0.95787    3186.6  0.134850  0.86515\n",
              "\n",
              "[5 rows x 20 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 108
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uGOSDVkVNNYA",
        "outputId": "c010c3bd-413e-4277-a6b6-76c5767f7d2a"
      },
      "source": [
        "X_fs.shape"
      ],
      "id": "uGOSDVkVNNYA",
      "execution_count": 109,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(43405, 20)"
            ]
          },
          "metadata": {},
          "execution_count": 36
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(43405, 20)"
            ]
          },
          "metadata": {},
          "execution_count": 109
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qlfIB922NQs3",
        "outputId": "33a32e2d-f2f2-4c0b-9247-75676eea8711"
      },
      "source": [
        "y.shape"
      ],
      "id": "qlfIB922NQs3",
      "execution_count": 110,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(43405,)"
            ]
          },
          "metadata": {},
          "execution_count": 37
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(43405,)"
            ]
          },
          "metadata": {},
          "execution_count": 110
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eFcZRbzzMcTi"
      },
      "source": [
        "# Stratify Trainn Test Split\n",
        "\n",
        "---\n",
        "\n",
        "test ratio =0.25 , stratify wrt to Y"
      ],
      "id": "eFcZRbzzMcTi"
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fjNJTUAiMbHl"
      },
      "source": [
        "# split into train test sets\n",
        "from sklearn.model_selection import train_test_split\n",
        "X_train, X_test, y_train, y_test = train_test_split(X_fs, y, test_size=0.25, random_state=1, stratify=y)"
      ],
      "id": "fjNJTUAiMbHl",
      "execution_count": 111,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 445
        },
        "id": "YgjvDIb8N3kC",
        "outputId": "a320e3f5-efe2-4c51-e63c-9994e2341df9"
      },
      "source": [
        "print(X_train.shape)\n",
        "X_train.head()"
      ],
      "id": "YgjvDIb8N3kC",
      "execution_count": 112,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(32553, 20)\n"
          ]
        },
        {
          "output_type": "execute_result",
          "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>Attr4</th>\n",
              "      <th>Attr5</th>\n",
              "      <th>Attr6</th>\n",
              "      <th>Attr13</th>\n",
              "      <th>Attr20</th>\n",
              "      <th>Attr22</th>\n",
              "      <th>Attr24</th>\n",
              "      <th>Attr25</th>\n",
              "      <th>Attr27</th>\n",
              "      <th>Attr29</th>\n",
              "      <th>Attr30</th>\n",
              "      <th>Attr34</th>\n",
              "      <th>Attr36</th>\n",
              "      <th>Attr39</th>\n",
              "      <th>Attr40</th>\n",
              "      <th>Attr44</th>\n",
              "      <th>Attr46</th>\n",
              "      <th>Attr55</th>\n",
              "      <th>Attr56</th>\n",
              "      <th>Attr58</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>10690</th>\n",
              "      <td>2.69590</td>\n",
              "      <td>58.91500</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.069412</td>\n",
              "      <td>0.91139</td>\n",
              "      <td>0.056731</td>\n",
              "      <td>0.039844</td>\n",
              "      <td>0.83127</td>\n",
              "      <td>1.486300</td>\n",
              "      <td>2.7209</td>\n",
              "      <td>0.017343</td>\n",
              "      <td>10.42000</td>\n",
              "      <td>0.81331</td>\n",
              "      <td>0.056373</td>\n",
              "      <td>0.80849</td>\n",
              "      <td>61.476</td>\n",
              "      <td>2.66840</td>\n",
              "      <td>65.690</td>\n",
              "      <td>0.056373</td>\n",
              "      <td>0.95287</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12771</th>\n",
              "      <td>3.31620</td>\n",
              "      <td>40.35500</td>\n",
              "      <td>0.120030</td>\n",
              "      <td>0.250290</td>\n",
              "      <td>20.75100</td>\n",
              "      <td>0.515000</td>\n",
              "      <td>1.313200</td>\n",
              "      <td>0.75656</td>\n",
              "      <td>123.680000</td>\n",
              "      <td>3.6308</td>\n",
              "      <td>0.047043</td>\n",
              "      <td>12.56500</td>\n",
              "      <td>2.38100</td>\n",
              "      <td>0.221620</td>\n",
              "      <td>0.29507</td>\n",
              "      <td>47.559</td>\n",
              "      <td>2.39840</td>\n",
              "      <td>1459.900</td>\n",
              "      <td>0.221620</td>\n",
              "      <td>0.78552</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11317</th>\n",
              "      <td>1.01100</td>\n",
              "      <td>-30.43600</td>\n",
              "      <td>-0.091448</td>\n",
              "      <td>0.028228</td>\n",
              "      <td>29.22500</td>\n",
              "      <td>0.039798</td>\n",
              "      <td>-0.107860</td>\n",
              "      <td>0.30894</td>\n",
              "      <td>0.276000</td>\n",
              "      <td>4.0965</td>\n",
              "      <td>0.431250</td>\n",
              "      <td>0.05759</td>\n",
              "      <td>1.49940</td>\n",
              "      <td>0.026672</td>\n",
              "      <td>0.11894</td>\n",
              "      <td>64.797</td>\n",
              "      <td>0.73369</td>\n",
              "      <td>59.000</td>\n",
              "      <td>0.033648</td>\n",
              "      <td>0.96635</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14216</th>\n",
              "      <td>0.49956</td>\n",
              "      <td>-1792.70000</td>\n",
              "      <td>0.000469</td>\n",
              "      <td>0.044002</td>\n",
              "      <td>14.93500</td>\n",
              "      <td>0.093398</td>\n",
              "      <td>0.064927</td>\n",
              "      <td>0.15086</td>\n",
              "      <td>2.129500</td>\n",
              "      <td>5.1265</td>\n",
              "      <td>0.506040</td>\n",
              "      <td>1.21380</td>\n",
              "      <td>1.16980</td>\n",
              "      <td>0.062900</td>\n",
              "      <td>0.19029</td>\n",
              "      <td>29.771</td>\n",
              "      <td>0.39624</td>\n",
              "      <td>-31019.000</td>\n",
              "      <td>0.300400</td>\n",
              "      <td>0.74140</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>34897</th>\n",
              "      <td>1.48640</td>\n",
              "      <td>-0.40519</td>\n",
              "      <td>-0.745230</td>\n",
              "      <td>0.025035</td>\n",
              "      <td>19.86800</td>\n",
              "      <td>0.000976</td>\n",
              "      <td>-0.896750</td>\n",
              "      <td>0.55472</td>\n",
              "      <td>1107.896314</td>\n",
              "      <td>2.8085</td>\n",
              "      <td>0.177570</td>\n",
              "      <td>3.54270</td>\n",
              "      <td>1.31770</td>\n",
              "      <td>0.011772</td>\n",
              "      <td>0.75702</td>\n",
              "      <td>27.400</td>\n",
              "      <td>1.17980</td>\n",
              "      <td>73.224</td>\n",
              "      <td>0.011772</td>\n",
              "      <td>0.99705</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "         Attr4       Attr5     Attr6  ...     Attr55    Attr56   Attr58\n",
              "10690  2.69590    58.91500  0.000000  ...     65.690  0.056373  0.95287\n",
              "12771  3.31620    40.35500  0.120030  ...   1459.900  0.221620  0.78552\n",
              "11317  1.01100   -30.43600 -0.091448  ...     59.000  0.033648  0.96635\n",
              "14216  0.49956 -1792.70000  0.000469  ... -31019.000  0.300400  0.74140\n",
              "34897  1.48640    -0.40519 -0.745230  ...     73.224  0.011772  0.99705\n",
              "\n",
              "[5 rows x 20 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 39
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(32553, 20)\n"
          ]
        },
        {
          "output_type": "execute_result",
          "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>Attr4</th>\n",
              "      <th>Attr5</th>\n",
              "      <th>Attr6</th>\n",
              "      <th>Attr13</th>\n",
              "      <th>Attr20</th>\n",
              "      <th>Attr22</th>\n",
              "      <th>Attr24</th>\n",
              "      <th>Attr25</th>\n",
              "      <th>Attr27</th>\n",
              "      <th>Attr29</th>\n",
              "      <th>Attr30</th>\n",
              "      <th>Attr34</th>\n",
              "      <th>Attr36</th>\n",
              "      <th>Attr39</th>\n",
              "      <th>Attr40</th>\n",
              "      <th>Attr44</th>\n",
              "      <th>Attr46</th>\n",
              "      <th>Attr55</th>\n",
              "      <th>Attr56</th>\n",
              "      <th>Attr58</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>10690</th>\n",
              "      <td>2.69590</td>\n",
              "      <td>58.91500</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.069412</td>\n",
              "      <td>0.91139</td>\n",
              "      <td>0.056731</td>\n",
              "      <td>0.039844</td>\n",
              "      <td>0.83127</td>\n",
              "      <td>1.486300</td>\n",
              "      <td>2.7209</td>\n",
              "      <td>0.017343</td>\n",
              "      <td>10.42000</td>\n",
              "      <td>0.81331</td>\n",
              "      <td>0.056373</td>\n",
              "      <td>0.80849</td>\n",
              "      <td>61.476</td>\n",
              "      <td>2.66840</td>\n",
              "      <td>65.690</td>\n",
              "      <td>0.056373</td>\n",
              "      <td>0.95287</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12771</th>\n",
              "      <td>3.31620</td>\n",
              "      <td>40.35500</td>\n",
              "      <td>0.120030</td>\n",
              "      <td>0.250290</td>\n",
              "      <td>20.75100</td>\n",
              "      <td>0.515000</td>\n",
              "      <td>1.313200</td>\n",
              "      <td>0.75656</td>\n",
              "      <td>123.680000</td>\n",
              "      <td>3.6308</td>\n",
              "      <td>0.047043</td>\n",
              "      <td>12.56500</td>\n",
              "      <td>2.38100</td>\n",
              "      <td>0.221620</td>\n",
              "      <td>0.29507</td>\n",
              "      <td>47.559</td>\n",
              "      <td>2.39840</td>\n",
              "      <td>1459.900</td>\n",
              "      <td>0.221620</td>\n",
              "      <td>0.78552</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11317</th>\n",
              "      <td>1.01100</td>\n",
              "      <td>-30.43600</td>\n",
              "      <td>-0.091448</td>\n",
              "      <td>0.028228</td>\n",
              "      <td>29.22500</td>\n",
              "      <td>0.039798</td>\n",
              "      <td>-0.107860</td>\n",
              "      <td>0.30894</td>\n",
              "      <td>0.276000</td>\n",
              "      <td>4.0965</td>\n",
              "      <td>0.431250</td>\n",
              "      <td>0.05759</td>\n",
              "      <td>1.49940</td>\n",
              "      <td>0.026672</td>\n",
              "      <td>0.11894</td>\n",
              "      <td>64.797</td>\n",
              "      <td>0.73369</td>\n",
              "      <td>59.000</td>\n",
              "      <td>0.033648</td>\n",
              "      <td>0.96635</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14216</th>\n",
              "      <td>0.49956</td>\n",
              "      <td>-1792.70000</td>\n",
              "      <td>0.000469</td>\n",
              "      <td>0.044002</td>\n",
              "      <td>14.93500</td>\n",
              "      <td>0.093398</td>\n",
              "      <td>0.064927</td>\n",
              "      <td>0.15086</td>\n",
              "      <td>2.129500</td>\n",
              "      <td>5.1265</td>\n",
              "      <td>0.506040</td>\n",
              "      <td>1.21380</td>\n",
              "      <td>1.16980</td>\n",
              "      <td>0.062900</td>\n",
              "      <td>0.19029</td>\n",
              "      <td>29.771</td>\n",
              "      <td>0.39624</td>\n",
              "      <td>-31019.000</td>\n",
              "      <td>0.300400</td>\n",
              "      <td>0.74140</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>34897</th>\n",
              "      <td>1.48640</td>\n",
              "      <td>-0.40519</td>\n",
              "      <td>-0.745230</td>\n",
              "      <td>0.025035</td>\n",
              "      <td>19.86800</td>\n",
              "      <td>0.000976</td>\n",
              "      <td>-0.896750</td>\n",
              "      <td>0.55472</td>\n",
              "      <td>1107.896314</td>\n",
              "      <td>2.8085</td>\n",
              "      <td>0.177570</td>\n",
              "      <td>3.54270</td>\n",
              "      <td>1.31770</td>\n",
              "      <td>0.011772</td>\n",
              "      <td>0.75702</td>\n",
              "      <td>27.400</td>\n",
              "      <td>1.17980</td>\n",
              "      <td>73.224</td>\n",
              "      <td>0.011772</td>\n",
              "      <td>0.99705</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "         Attr4       Attr5     Attr6  ...     Attr55    Attr56   Attr58\n",
              "10690  2.69590    58.91500  0.000000  ...     65.690  0.056373  0.95287\n",
              "12771  3.31620    40.35500  0.120030  ...   1459.900  0.221620  0.78552\n",
              "11317  1.01100   -30.43600 -0.091448  ...     59.000  0.033648  0.96635\n",
              "14216  0.49956 -1792.70000  0.000469  ... -31019.000  0.300400  0.74140\n",
              "34897  1.48640    -0.40519 -0.745230  ...     73.224  0.011772  0.99705\n",
              "\n",
              "[5 rows x 20 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 112
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 445
        },
        "id": "neQSw6zcOBjC",
        "outputId": "f74f0ffb-fb96-4961-f3a6-993ef6c17f6c"
      },
      "source": [
        "print(X_test.shape)\n",
        "X_test.head()"
      ],
      "id": "neQSw6zcOBjC",
      "execution_count": 113,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(10852, 20)\n"
          ]
        },
        {
          "output_type": "execute_result",
          "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>Attr4</th>\n",
              "      <th>Attr5</th>\n",
              "      <th>Attr6</th>\n",
              "      <th>Attr13</th>\n",
              "      <th>Attr20</th>\n",
              "      <th>Attr22</th>\n",
              "      <th>Attr24</th>\n",
              "      <th>Attr25</th>\n",
              "      <th>Attr27</th>\n",
              "      <th>Attr29</th>\n",
              "      <th>Attr30</th>\n",
              "      <th>Attr34</th>\n",
              "      <th>Attr36</th>\n",
              "      <th>Attr39</th>\n",
              "      <th>Attr40</th>\n",
              "      <th>Attr44</th>\n",
              "      <th>Attr46</th>\n",
              "      <th>Attr55</th>\n",
              "      <th>Attr56</th>\n",
              "      <th>Attr58</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>28232</th>\n",
              "      <td>4.4101</td>\n",
              "      <td>84.501</td>\n",
              "      <td>0.068618</td>\n",
              "      <td>0.146040</td>\n",
              "      <td>45.297</td>\n",
              "      <td>0.283820</td>\n",
              "      <td>0.575280</td>\n",
              "      <td>0.73753</td>\n",
              "      <td>41.82200</td>\n",
              "      <td>4.3740</td>\n",
              "      <td>0.05416</td>\n",
              "      <td>7.747300</td>\n",
              "      <td>2.00860</td>\n",
              "      <td>0.142480</td>\n",
              "      <td>0.926020</td>\n",
              "      <td>91.520</td>\n",
              "      <td>3.25660</td>\n",
              "      <td>17435.000</td>\n",
              "      <td>0.142480</td>\n",
              "      <td>0.86217</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>18108</th>\n",
              "      <td>1.2467</td>\n",
              "      <td>-48.957</td>\n",
              "      <td>0.094732</td>\n",
              "      <td>0.092668</td>\n",
              "      <td>69.381</td>\n",
              "      <td>0.026318</td>\n",
              "      <td>0.119420</td>\n",
              "      <td>0.39481</td>\n",
              "      <td>0.31803</td>\n",
              "      <td>4.3291</td>\n",
              "      <td>0.58504</td>\n",
              "      <td>0.050604</td>\n",
              "      <td>0.88423</td>\n",
              "      <td>0.030996</td>\n",
              "      <td>0.101290</td>\n",
              "      <td>44.352</td>\n",
              "      <td>0.54796</td>\n",
              "      <td>1215.600</td>\n",
              "      <td>0.025358</td>\n",
              "      <td>0.97464</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11653</th>\n",
              "      <td>1.1721</td>\n",
              "      <td>-93.484</td>\n",
              "      <td>0.069834</td>\n",
              "      <td>0.059641</td>\n",
              "      <td>107.130</td>\n",
              "      <td>0.138270</td>\n",
              "      <td>0.094746</td>\n",
              "      <td>0.28626</td>\n",
              "      <td>0.73030</td>\n",
              "      <td>4.0834</td>\n",
              "      <td>0.30321</td>\n",
              "      <td>0.193730</td>\n",
              "      <td>2.06210</td>\n",
              "      <td>0.067868</td>\n",
              "      <td>0.140290</td>\n",
              "      <td>20.003</td>\n",
              "      <td>0.30263</td>\n",
              "      <td>1434.200</td>\n",
              "      <td>0.070680</td>\n",
              "      <td>0.92932</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19879</th>\n",
              "      <td>1.0264</td>\n",
              "      <td>-74.500</td>\n",
              "      <td>0.014910</td>\n",
              "      <td>0.132190</td>\n",
              "      <td>55.612</td>\n",
              "      <td>0.129830</td>\n",
              "      <td>0.282760</td>\n",
              "      <td>0.29853</td>\n",
              "      <td>1.31880</td>\n",
              "      <td>4.2299</td>\n",
              "      <td>0.81509</td>\n",
              "      <td>1.032400</td>\n",
              "      <td>0.85350</td>\n",
              "      <td>0.157370</td>\n",
              "      <td>0.046333</td>\n",
              "      <td>11.155</td>\n",
              "      <td>0.21008</td>\n",
              "      <td>71.464</td>\n",
              "      <td>0.157370</td>\n",
              "      <td>0.96280</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6169</th>\n",
              "      <td>1.0165</td>\n",
              "      <td>-41.710</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.057796</td>\n",
              "      <td>39.429</td>\n",
              "      <td>0.109070</td>\n",
              "      <td>0.115150</td>\n",
              "      <td>0.16604</td>\n",
              "      <td>3.26970</td>\n",
              "      <td>4.0222</td>\n",
              "      <td>0.35326</td>\n",
              "      <td>2.547100</td>\n",
              "      <td>2.21720</td>\n",
              "      <td>0.047440</td>\n",
              "      <td>0.088880</td>\n",
              "      <td>56.407</td>\n",
              "      <td>0.63484</td>\n",
              "      <td>108.770</td>\n",
              "      <td>0.047440</td>\n",
              "      <td>0.96449</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "        Attr4   Attr5     Attr6  ...     Attr55    Attr56   Attr58\n",
              "28232  4.4101  84.501  0.068618  ...  17435.000  0.142480  0.86217\n",
              "18108  1.2467 -48.957  0.094732  ...   1215.600  0.025358  0.97464\n",
              "11653  1.1721 -93.484  0.069834  ...   1434.200  0.070680  0.92932\n",
              "19879  1.0264 -74.500  0.014910  ...     71.464  0.157370  0.96280\n",
              "6169   1.0165 -41.710  0.000000  ...    108.770  0.047440  0.96449\n",
              "\n",
              "[5 rows x 20 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 40
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(10852, 20)\n"
          ]
        },
        {
          "output_type": "execute_result",
          "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>Attr4</th>\n",
              "      <th>Attr5</th>\n",
              "      <th>Attr6</th>\n",
              "      <th>Attr13</th>\n",
              "      <th>Attr20</th>\n",
              "      <th>Attr22</th>\n",
              "      <th>Attr24</th>\n",
              "      <th>Attr25</th>\n",
              "      <th>Attr27</th>\n",
              "      <th>Attr29</th>\n",
              "      <th>Attr30</th>\n",
              "      <th>Attr34</th>\n",
              "      <th>Attr36</th>\n",
              "      <th>Attr39</th>\n",
              "      <th>Attr40</th>\n",
              "      <th>Attr44</th>\n",
              "      <th>Attr46</th>\n",
              "      <th>Attr55</th>\n",
              "      <th>Attr56</th>\n",
              "      <th>Attr58</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>28232</th>\n",
              "      <td>4.4101</td>\n",
              "      <td>84.501</td>\n",
              "      <td>0.068618</td>\n",
              "      <td>0.146040</td>\n",
              "      <td>45.297</td>\n",
              "      <td>0.283820</td>\n",
              "      <td>0.575280</td>\n",
              "      <td>0.73753</td>\n",
              "      <td>41.82200</td>\n",
              "      <td>4.3740</td>\n",
              "      <td>0.05416</td>\n",
              "      <td>7.747300</td>\n",
              "      <td>2.00860</td>\n",
              "      <td>0.142480</td>\n",
              "      <td>0.926020</td>\n",
              "      <td>91.520</td>\n",
              "      <td>3.25660</td>\n",
              "      <td>17435.000</td>\n",
              "      <td>0.142480</td>\n",
              "      <td>0.86217</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>18108</th>\n",
              "      <td>1.2467</td>\n",
              "      <td>-48.957</td>\n",
              "      <td>0.094732</td>\n",
              "      <td>0.092668</td>\n",
              "      <td>69.381</td>\n",
              "      <td>0.026318</td>\n",
              "      <td>0.119420</td>\n",
              "      <td>0.39481</td>\n",
              "      <td>0.31803</td>\n",
              "      <td>4.3291</td>\n",
              "      <td>0.58504</td>\n",
              "      <td>0.050604</td>\n",
              "      <td>0.88423</td>\n",
              "      <td>0.030996</td>\n",
              "      <td>0.101290</td>\n",
              "      <td>44.352</td>\n",
              "      <td>0.54796</td>\n",
              "      <td>1215.600</td>\n",
              "      <td>0.025358</td>\n",
              "      <td>0.97464</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11653</th>\n",
              "      <td>1.1721</td>\n",
              "      <td>-93.484</td>\n",
              "      <td>0.069834</td>\n",
              "      <td>0.059641</td>\n",
              "      <td>107.130</td>\n",
              "      <td>0.138270</td>\n",
              "      <td>0.094746</td>\n",
              "      <td>0.28626</td>\n",
              "      <td>0.73030</td>\n",
              "      <td>4.0834</td>\n",
              "      <td>0.30321</td>\n",
              "      <td>0.193730</td>\n",
              "      <td>2.06210</td>\n",
              "      <td>0.067868</td>\n",
              "      <td>0.140290</td>\n",
              "      <td>20.003</td>\n",
              "      <td>0.30263</td>\n",
              "      <td>1434.200</td>\n",
              "      <td>0.070680</td>\n",
              "      <td>0.92932</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19879</th>\n",
              "      <td>1.0264</td>\n",
              "      <td>-74.500</td>\n",
              "      <td>0.014910</td>\n",
              "      <td>0.132190</td>\n",
              "      <td>55.612</td>\n",
              "      <td>0.129830</td>\n",
              "      <td>0.282760</td>\n",
              "      <td>0.29853</td>\n",
              "      <td>1.31880</td>\n",
              "      <td>4.2299</td>\n",
              "      <td>0.81509</td>\n",
              "      <td>1.032400</td>\n",
              "      <td>0.85350</td>\n",
              "      <td>0.157370</td>\n",
              "      <td>0.046333</td>\n",
              "      <td>11.155</td>\n",
              "      <td>0.21008</td>\n",
              "      <td>71.464</td>\n",
              "      <td>0.157370</td>\n",
              "      <td>0.96280</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6169</th>\n",
              "      <td>1.0165</td>\n",
              "      <td>-41.710</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.057796</td>\n",
              "      <td>39.429</td>\n",
              "      <td>0.109070</td>\n",
              "      <td>0.115150</td>\n",
              "      <td>0.16604</td>\n",
              "      <td>3.26970</td>\n",
              "      <td>4.0222</td>\n",
              "      <td>0.35326</td>\n",
              "      <td>2.547100</td>\n",
              "      <td>2.21720</td>\n",
              "      <td>0.047440</td>\n",
              "      <td>0.088880</td>\n",
              "      <td>56.407</td>\n",
              "      <td>0.63484</td>\n",
              "      <td>108.770</td>\n",
              "      <td>0.047440</td>\n",
              "      <td>0.96449</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "        Attr4   Attr5     Attr6  ...     Attr55    Attr56   Attr58\n",
              "28232  4.4101  84.501  0.068618  ...  17435.000  0.142480  0.86217\n",
              "18108  1.2467 -48.957  0.094732  ...   1215.600  0.025358  0.97464\n",
              "11653  1.1721 -93.484  0.069834  ...   1434.200  0.070680  0.92932\n",
              "19879  1.0264 -74.500  0.014910  ...     71.464  0.157370  0.96280\n",
              "6169   1.0165 -41.710  0.000000  ...    108.770  0.047440  0.96449\n",
              "\n",
              "[5 rows x 20 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 113
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NjxjWNe1A7hc",
        "outputId": "eb83d6ef-bfae-47a9-f7dc-71b64d44de53"
      },
      "source": [
        "y_train.value_counts()\n",
        "nzero=y_train.value_counts()[0];\n",
        "n1= y_train.value_counts()[1]\n",
        "ratio_train = n1/nzero\n",
        "print('ratio_train=',ratio_train)"
      ],
      "id": "NjxjWNe1A7hc",
      "execution_count": 114,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "ratio_train= 0.05060513151524931\n",
            "ratio_train= 0.05060513151524931\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1B48dixUBGtL",
        "outputId": "7260332b-ef54-423f-ba1e-9331a32689d9"
      },
      "source": [
        "y_test.value_counts()\n",
        "y_train.value_counts()\n",
        "nzero=y_test.value_counts()[0];\n",
        "n1= y_test.value_counts()[1]\n",
        "ratio_test = n1/nzero\n",
        "print('ratio_test=',ratio_test)\n",
        "print()\n",
        "print('difference in ratio : train-test=',int(np.abs(ratio_train-ratio_test)))"
      ],
      "id": "1B48dixUBGtL",
      "execution_count": 115,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "ratio_test= 0.050634136896117725\n",
            "\n",
            "difference in ratio : train-test= 0\n",
            "ratio_test= 0.050634136896117725\n",
            "\n",
            "difference in ratio : train-test= 0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XX0fsl4U6dZY"
      },
      "source": [
        "# Model 1: Logistic Regression \n",
        "\n",
        "---\n",
        "\n"
      ],
      "id": "XX0fsl4U6dZY"
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 587
        },
        "id": "AIjh6lKeOvD0",
        "outputId": "264b3232-58fa-4730-8275-29cc757b7944"
      },
      "source": [
        "sns.countplot(y_train)"
      ],
      "id": "AIjh6lKeOvD0",
      "execution_count": 116,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f07822e5650>"
            ]
          },
          "metadata": {},
          "execution_count": 43
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f07809fbc10>"
            ]
          },
          "metadata": {},
          "execution_count": 116
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "e8vdep4FO0d2"
      },
      "source": [
        "**#imbalanced Data stratified test-train data**"
      ],
      "id": "e8vdep4FO0d2"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TXWSHNTFas6x"
      },
      "source": [
        "\n",
        "\n",
        "\n",
        "```\n",
        "##hyperparameter tuning\n",
        "#Grid Search\n",
        "logreg = LogisticRegression(penalty='l2')\n",
        "param = {'C':[0.001,0.003,0.005,0.01,0.03,0.05,0.1,0.3,0.5,1,2,3,3,4,5,10,20]}\n",
        "clf = GridSearchCV(logreg,param)\n",
        "clf.fit(X_train,y_train)\n",
        "print('Best roc_auc: {:.4}, with best C: {}'.format(clf.best_score_, clf.best_params_))\n",
        "```\n",
        "\n",
        "best C: {'C': 0.001}"
      ],
      "id": "TXWSHNTFas6x"
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nlnhHQUZOdwL",
        "outputId": "ea7d411d-638c-49f0-e8c0-b95401e1af34"
      },
      "source": [
        "model = LogisticRegression(random_state=42,penalty='l2',C=0.001)\n",
        "\n",
        "model.fit(X_train, y_train)\n",
        "\n",
        "y_pred = model.predict(X_test)\n",
        "\n",
        "from sklearn.metrics import accuracy_score, classification_report\n",
        "\n",
        "print(\"Accuracy : \",round(accuracy_score(y_test, y_pred)*100,2))\n",
        "print(\" Train Accuracy : \",round(accuracy_score(y_train, model.predict(X_train))*100,2))"
      ],
      "id": "nlnhHQUZOdwL",
      "execution_count": 117,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy :  94.66\n",
            " Train Accuracy :  94.57\n",
            "Accuracy :  94.66\n",
            " Train Accuracy :  94.57\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ROdH6LZjMoT4",
        "outputId": "68a1a4f6-a619-452f-ad1f-0c162b7151f5"
      },
      "source": [
        "\n",
        "kf = StratifiedKFold(n_splits=5,shuffle=True)\n",
        "pred_test_full =0\n",
        "cv_score =[]\n",
        "i=1\n",
        "for train_index,test_index in kf.split(X_train,y_train):\n",
        "    print('{} of KFold {}'.format(i,kf.n_splits))\n",
        "    xtr,xvl = X_train.iloc[train_index],X_train.iloc[test_index]\n",
        "    ytr,yvl = y_train.iloc[train_index],y_train.iloc[test_index]\n",
        "    \n",
        "    #model\n",
        "    lr = LogisticRegression(penalty='l2')\n",
        "    lr.fit(xtr,ytr)\n",
        "    score = accuracy_score(yvl,lr.predict(xvl))\n",
        "    print('Accuracy Score:',score)\n",
        "    cv_score.append(score)    \n",
        "    pred_test = lr.predict_proba(X_test)[:,1]\n",
        "    pred_test_full +=pred_test    \n",
        "    i+=1"
      ],
      "id": "ROdH6LZjMoT4",
      "execution_count": 118,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1 of KFold 5\n",
            "Accuracy Score: 0.9516203348180003\n",
            "2 of KFold 5\n",
            "Accuracy Score: 0.9493165412379051\n",
            "3 of KFold 5\n",
            "Accuracy Score: 0.9513131623406543\n",
            "4 of KFold 5\n",
            "Accuracy Score: 0.9508448540706606\n",
            "5 of KFold 5\n",
            "Accuracy Score: 0.9513056835637481\n",
            "1 of KFold 5\n",
            "Accuracy Score: 0.9506988173859622\n",
            "2 of KFold 5\n",
            "Accuracy Score: 0.9510059898633082\n",
            "3 of KFold 5\n",
            "Accuracy Score: 0.9510059898633082\n",
            "4 of KFold 5\n",
            "Accuracy Score: 0.9499231950844854\n",
            "5 of KFold 5\n",
            "Accuracy Score: 0.9513056835637481\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YVp0gT1A0Rss",
        "outputId": "374ea5c5-c336-42ee-acd5-2089c734c874"
      },
      "source": [
        "from sklearn import metrics\n",
        "cm = metrics.confusion_matrix(y_test, model.predict(X_test))\n",
        "print(cm)"
      ],
      "id": "YVp0gT1A0Rss",
      "execution_count": 119,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[10268    61]\n",
            " [  519     4]]\n",
            "[[10268    61]\n",
            " [  519     4]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 711
        },
        "id": "f5hBXWvf5s92",
        "outputId": "0800f6c6-08fb-4736-dfcd-5deb5c88f855"
      },
      "source": [
        "plt.figure(figsize=(6,6))\n",
        "sns.heatmap((cm/np.sum(cm)*100), annot=True, fmt=\".3f\", linewidths=.5, square = True, cmap = 'viridis');\n",
        "plt.ylabel('Actual label');\n",
        "plt.xlabel('Predicted label');\n",
        "all_sample_title = 'Accuracy Score: {0}'.format(accuracy_score(y_train, model.predict(X_train))*100)\n",
        "plt.title(all_sample_title, size = 15);"
      ],
      "id": "f5hBXWvf5s92",
      "execution_count": 120,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x432 with 2 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x432 with 2 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 143
        },
        "id": "X-yl6XUXRsmC",
        "outputId": "0d565a86-114b-4754-f2d6-dc4d61ecf306"
      },
      "source": [
        "from sklearn.metrics import accuracy_score, f1_score,recall_score,precision_score, confusion_matrix\n",
        "acc = accuracy_score(y_test, y_pred)\n",
        "prec = precision_score(y_test, y_pred)\n",
        "rec = recall_score(y_test, y_pred)\n",
        "f1 = f1_score(y_test, y_pred)\n",
        "results = pd.DataFrame([['Logistic Regression ', acc*100,prec*100,rec*100,f1*100]],\n",
        "                       columns=['Model', 'Accuracy', 'Precision', 'Recall','F1 Score'])\n",
        "results"
      ],
      "id": "X-yl6XUXRsmC",
      "execution_count": 121,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>Model</th>\n",
              "      <th>Accuracy</th>\n",
              "      <th>Precision</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1 Score</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Logistic Regression</td>\n",
              "      <td>94.655363</td>\n",
              "      <td>6.153846</td>\n",
              "      <td>0.764818</td>\n",
              "      <td>1.360544</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                  Model   Accuracy  Precision    Recall  F1 Score\n",
              "0  Logistic Regression   94.655363   6.153846  0.764818  1.360544"
            ]
          },
          "metadata": {},
          "execution_count": 48
        },
        {
          "output_type": "execute_result",
          "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>Model</th>\n",
              "      <th>Accuracy</th>\n",
              "      <th>Precision</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1 Score</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Logistic Regression</td>\n",
              "      <td>94.655363</td>\n",
              "      <td>6.153846</td>\n",
              "      <td>0.764818</td>\n",
              "      <td>1.360544</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                  Model   Accuracy  Precision    Recall  F1 Score\n",
              "0  Logistic Regression   94.655363   6.153846  0.764818  1.360544"
            ]
          },
          "metadata": {},
          "execution_count": 121
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4lcR5mBA0SR7"
      },
      "source": [
        "# Model 2: Neural Network--> Scikit Learn\n",
        "\n",
        "---\n",
        "\n"
      ],
      "id": "4lcR5mBA0SR7"
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "shZg7nSuSMcj",
        "outputId": "63f86b0a-64f4-409d-82df-05fca72b1f8f"
      },
      "source": [
        "from sklearn.neural_network import MLPClassifier\n",
        "from sklearn.metrics  import classification_report\n",
        "model2 = MLPClassifier(random_state=42)\n",
        "\n",
        "model2.fit(X_train, y_train)\n",
        "\n",
        "y_pred2 = model2.predict(X_test)\n",
        "\n",
        "from sklearn.metrics import accuracy_score, classification_report\n",
        "\n",
        "print(\"Accuracy : \",round(accuracy_score(y_test, y_pred2)*100,2))"
      ],
      "id": "shZg7nSuSMcj",
      "execution_count": 122,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy :  92.53\n",
            "Accuracy :  92.53\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vZO9a-BZSV5j",
        "outputId": "13a82edf-6103-48e8-e5cf-a2f6dbc0cfca"
      },
      "source": [
        "print(classification_report(y_test, y_pred2))"
      ],
      "id": "vZO9a-BZSV5j",
      "execution_count": 123,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.96      0.96      0.96     10329\n",
            "           1       0.21      0.20      0.20       523\n",
            "\n",
            "    accuracy                           0.93     10852\n",
            "   macro avg       0.58      0.58      0.58     10852\n",
            "weighted avg       0.92      0.93      0.92     10852\n",
            "\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.96      0.96      0.96     10329\n",
            "           1       0.21      0.20      0.20       523\n",
            "\n",
            "    accuracy                           0.93     10852\n",
            "   macro avg       0.58      0.58      0.58     10852\n",
            "weighted avg       0.92      0.93      0.92     10852\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 779
        },
        "id": "YnEoG0V7TDKI",
        "outputId": "7bb86c7c-1d9a-401e-d5e6-c6f449fe9cc7"
      },
      "source": [
        "from sklearn import metrics\n",
        "cm = metrics.confusion_matrix(y_test, y_pred2)\n",
        "print(cm)\n",
        "\n",
        "plt.figure(figsize=(6,6))\n",
        "sns.heatmap((cm/np.sum(cm)*100), annot=True, fmt=\".3f\", linewidths=.5, square = True, cmap = 'viridis');\n",
        "plt.ylabel('Actual label');\n",
        "plt.xlabel('Predicted label');\n",
        "all_sample_title = 'Accuracy Score: {0}'.format(accuracy_score(y_test, y_pred2)*100)\n",
        "plt.title(all_sample_title, size = 15);"
      ],
      "id": "YnEoG0V7TDKI",
      "execution_count": 124,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[9937  392]\n",
            " [ 419  104]]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x432 with 2 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[9937  392]\n",
            " [ 419  104]]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x432 with 2 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LgoY-1LeZI83"
      },
      "source": [
        "from sklearn.metrics import accuracy_score, f1_score,recall_score,precision_score, confusion_matrix\n",
        "acc = accuracy_score(y_test, y_pred2)\n",
        "prec = precision_score(y_test, y_pred2)\n",
        "rec = recall_score(y_test, y_pred2)\n",
        "f1 = f1_score(y_test, y_pred2)\n",
        "\n",
        "results.loc[len(results.index)] =['MLP Classifier', acc*100,\n",
        "                                  prec*100,rec*100,f1*100]"
      ],
      "id": "LgoY-1LeZI83",
      "execution_count": 125,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gFsWFhnK8-2F"
      },
      "source": [
        "#  Model 3: Bernoulli Naive Bayes Classifier\n",
        "\n",
        "---\n",
        "\n"
      ],
      "id": "gFsWFhnK8-2F"
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EmRESrtg9Fj4",
        "outputId": "1a04b33c-583b-4323-8cfd-d63b9ca0a12a"
      },
      "source": [
        "from sklearn.naive_bayes import BernoulliNB\n",
        "model3 = BernoulliNB()\n",
        "model3.fit(X_train, y_train)\n",
        "\n",
        "y_pred2 = model3.predict(X_test)\n",
        "from sklearn.metrics import accuracy_score, classification_report\n",
        "print(\"Accuracy : \",round(accuracy_score(y_test, y_pred2)*100,2))\n",
        "print(classification_report(y_test, y_pred2))"
      ],
      "id": "EmRESrtg9Fj4",
      "execution_count": 126,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy :  86.33\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.96      0.89      0.93     10329\n",
            "           1       0.14      0.36      0.20       523\n",
            "\n",
            "    accuracy                           0.86     10852\n",
            "   macro avg       0.55      0.63      0.56     10852\n",
            "weighted avg       0.93      0.86      0.89     10852\n",
            "\n",
            "Accuracy :  86.33\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.96      0.89      0.93     10329\n",
            "           1       0.14      0.36      0.20       523\n",
            "\n",
            "    accuracy                           0.86     10852\n",
            "   macro avg       0.55      0.63      0.56     10852\n",
            "weighted avg       0.93      0.86      0.89     10852\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 681
        },
        "id": "z9o0jexB9FdW",
        "outputId": "3cf0cb85-0d93-4bea-e3a7-12e3b374a052"
      },
      "source": [
        "from sklearn import metrics\n",
        "cm = metrics.confusion_matrix(y_test, y_pred2)\n",
        "print(cm)\n",
        "\n",
        "plt.figure(figsize=(5,5))\n",
        "sns.heatmap((cm/np.sum(cm)*100), annot=True, fmt=\".3f\", linewidths=.5, square = True, cmap = 'viridis');\n",
        "plt.ylabel('Actual label');\n",
        "plt.xlabel('Predicted label');\n",
        "all_sample_title = 'Accuracy Score: {0}'.format(accuracy_score(y_test, y_pred2)*100)\n",
        "plt.title(all_sample_title, size = 15);"
      ],
      "id": "z9o0jexB9FdW",
      "execution_count": 127,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[9179 1150]\n",
            " [ 333  190]]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x360 with 2 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[9179 1150]\n",
            " [ 333  190]]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x360 with 2 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_qCOxkqMZ5L5"
      },
      "source": [
        "acc = accuracy_score(y_test, y_pred2)\n",
        "prec = precision_score(y_test, y_pred2)\n",
        "rec = recall_score(y_test, y_pred2)\n",
        "f1 = f1_score(y_test, y_pred2)\n",
        "results.loc[len(results.index)] =['BNB classifier', acc*100,\n",
        "                                  prec*100,rec*100,f1*100]"
      ],
      "id": "_qCOxkqMZ5L5",
      "execution_count": 128,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_EoZkSK5C-Er"
      },
      "source": [
        "\n",
        "\n",
        "---\n",
        "\n",
        "# Model 4: LGBM Classifier"
      ],
      "id": "_EoZkSK5C-Er"
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lwcNDJq_DNwR",
        "outputId": "055c9a5a-530e-45c5-d880-5dfbf70c1ce1"
      },
      "source": [
        "import lightgbm as lgb\n",
        "model4 = lgb.LGBMClassifier()\n",
        "model4.fit(X_train, y_train)\n",
        "\n",
        "y_pred2 = model4.predict(X_test)\n",
        "from sklearn.metrics import accuracy_score, classification_report\n",
        "print(\"Accuracy : \",round(accuracy_score(y_test, y_pred2)*100,2))\n",
        "print(classification_report(y_test, y_pred2))"
      ],
      "id": "lwcNDJq_DNwR",
      "execution_count": 129,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy :  97.59\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.98      1.00      0.99     10329\n",
            "           1       0.95      0.53      0.68       523\n",
            "\n",
            "    accuracy                           0.98     10852\n",
            "   macro avg       0.96      0.76      0.83     10852\n",
            "weighted avg       0.98      0.98      0.97     10852\n",
            "\n",
            "Accuracy :  97.59\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.98      1.00      0.99     10329\n",
            "           1       0.95      0.53      0.68       523\n",
            "\n",
            "    accuracy                           0.98     10852\n",
            "   macro avg       0.96      0.76      0.83     10852\n",
            "weighted avg       0.98      0.98      0.97     10852\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 681
        },
        "id": "9RhDBAQlDXcX",
        "outputId": "85c47d8b-a35a-4ada-95e7-56e0e8e698a7"
      },
      "source": [
        "from sklearn import metrics\n",
        "cm = metrics.confusion_matrix(y_test, y_pred2)\n",
        "print(cm)\n",
        "\n",
        "plt.figure(figsize=(5,5))\n",
        "sns.heatmap((cm/np.sum(cm)*100), annot=True, fmt=\".3f\", linewidths=.5, square = True, cmap = 'viridis');\n",
        "plt.ylabel('Actual label');\n",
        "plt.xlabel('Predicted label');\n",
        "all_sample_title = 'Accuracy Score: {0}'.format(accuracy_score(y_test, y_pred2)*100)\n",
        "plt.title(all_sample_title, size = 15);"
      ],
      "id": "9RhDBAQlDXcX",
      "execution_count": 130,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[10315    14]\n",
            " [  247   276]]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x360 with 2 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[10315    14]\n",
            " [  247   276]]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x360 with 2 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bmmm7OCbaOco"
      },
      "source": [
        "acc = accuracy_score(y_test, y_pred2)\n",
        "prec = precision_score(y_test, y_pred2)\n",
        "rec = recall_score(y_test, y_pred2)\n",
        "f1 = f1_score(y_test, y_pred2)\n",
        "results.loc[len(results.index)] =['LGBM classifier', acc*100,\n",
        "                                  prec*100,rec*100,f1*100]"
      ],
      "id": "bmmm7OCbaOco",
      "execution_count": 131,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sJQ2OZbfHHU5"
      },
      "source": [
        "\n",
        "\n",
        "---\n",
        "\n",
        "\n",
        "# Model 5: AdaBoost Classifier"
      ],
      "id": "sJQ2OZbfHHU5"
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "b5CBx6ZHHU5M",
        "outputId": "efc6cca8-11b6-4e70-ed82-09049d179264"
      },
      "source": [
        "from sklearn.ensemble import AdaBoostClassifier\n",
        "model5 = AdaBoostClassifier()\n",
        "model5.fit(X_train, y_train)\n",
        "\n",
        "y_pred2 = model5.predict(X_test)\n",
        "from sklearn.metrics import accuracy_score, classification_report\n",
        "print(\"Accuracy : \",round(accuracy_score(y_test, y_pred2)*100,2))\n",
        "print(classification_report(y_test, y_pred2))"
      ],
      "id": "b5CBx6ZHHU5M",
      "execution_count": 132,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy :  95.46\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.96      0.99      0.98     10329\n",
            "           1       0.59      0.19      0.29       523\n",
            "\n",
            "    accuracy                           0.95     10852\n",
            "   macro avg       0.77      0.59      0.63     10852\n",
            "weighted avg       0.94      0.95      0.94     10852\n",
            "\n",
            "Accuracy :  95.46\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.96      0.99      0.98     10329\n",
            "           1       0.59      0.19      0.29       523\n",
            "\n",
            "    accuracy                           0.95     10852\n",
            "   macro avg       0.77      0.59      0.63     10852\n",
            "weighted avg       0.94      0.95      0.94     10852\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 681
        },
        "id": "2Q4_QKOBH1Qf",
        "outputId": "b24c5d0a-f510-4558-b276-4a33a10d5efc"
      },
      "source": [
        "from sklearn import metrics\n",
        "cm = metrics.confusion_matrix(y_test, y_pred2)\n",
        "print(cm)\n",
        "\n",
        "plt.figure(figsize=(5,5))\n",
        "sns.heatmap((cm/np.sum(cm)*100), annot=True, fmt=\".3f\", linewidths=.5, square = True, cmap = 'viridis');\n",
        "plt.ylabel('Actual label');\n",
        "plt.xlabel('Predicted label');\n",
        "all_sample_title = 'Accuracy Score: {0}'.format(accuracy_score(y_test, y_pred2)*100)\n",
        "plt.title(all_sample_title, size = 15);"
      ],
      "id": "2Q4_QKOBH1Qf",
      "execution_count": 133,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[10258    71]\n",
            " [  422   101]]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x360 with 2 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[10258    71]\n",
            " [  422   101]]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x360 with 2 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Nh4VfkkKaVZX"
      },
      "source": [
        "acc = accuracy_score(y_test, y_pred2)\n",
        "prec = precision_score(y_test, y_pred2)\n",
        "rec = recall_score(y_test, y_pred2)\n",
        "f1 = f1_score(y_test, y_pred2)\n",
        "results.loc[len(results.index)] =['AdaBoost', acc*100,\n",
        "                                  prec*100,rec*100,f1*100]"
      ],
      "id": "Nh4VfkkKaVZX",
      "execution_count": 134,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_3PTF6_pVQsK"
      },
      "source": [
        "\n",
        "\n",
        "---\n",
        "\n",
        "\n",
        "# Summary Table - Without SMOTE"
      ],
      "id": "_3PTF6_pVQsK"
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        },
        "id": "L3RC_JO2aWdG",
        "outputId": "fc323d59-9da6-49b1-da31-194afc60e306"
      },
      "source": [
        "results.head(8)"
      ],
      "id": "L3RC_JO2aWdG",
      "execution_count": 146,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>Model</th>\n",
              "      <th>Accuracy</th>\n",
              "      <th>Precision</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1 Score</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Logistic Regression</td>\n",
              "      <td>94.655363</td>\n",
              "      <td>6.153846</td>\n",
              "      <td>0.764818</td>\n",
              "      <td>1.360544</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>MLP Classifier</td>\n",
              "      <td>92.526723</td>\n",
              "      <td>20.967742</td>\n",
              "      <td>19.885277</td>\n",
              "      <td>20.412169</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>BNB classifier</td>\n",
              "      <td>86.334316</td>\n",
              "      <td>14.179104</td>\n",
              "      <td>36.328872</td>\n",
              "      <td>20.397209</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>LGBM classifier</td>\n",
              "      <td>97.594913</td>\n",
              "      <td>95.172414</td>\n",
              "      <td>52.772467</td>\n",
              "      <td>67.896679</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>AdaBoost</td>\n",
              "      <td>95.457059</td>\n",
              "      <td>58.720930</td>\n",
              "      <td>19.311663</td>\n",
              "      <td>29.064748</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                  Model   Accuracy  Precision     Recall   F1 Score\n",
              "0  Logistic Regression   94.655363   6.153846   0.764818   1.360544\n",
              "1        MLP Classifier  92.526723  20.967742  19.885277  20.412169\n",
              "2        BNB classifier  86.334316  14.179104  36.328872  20.397209\n",
              "3       LGBM classifier  97.594913  95.172414  52.772467  67.896679\n",
              "4              AdaBoost  95.457059  58.720930  19.311663  29.064748"
            ]
          },
          "metadata": {},
          "execution_count": 146
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lfpsqyq59EBf"
      },
      "source": [
        "\n",
        "\n",
        "---\n",
        "\n",
        "\n",
        "\n",
        "---\n",
        "\n"
      ],
      "id": "lfpsqyq59EBf"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gqkjGLHrCtaS"
      },
      "source": [
        "\n",
        "\n",
        "\n",
        "\n",
        "# Checking on applying smote to train test data*\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ],
      "id": "gqkjGLHrCtaS"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KnF2MmeV9Ks3"
      },
      "source": [
        "#### ` Smote : Model 2`"
      ],
      "id": "KnF2MmeV9Ks3"
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GjAC0CFa9Y7m",
        "outputId": "ca3374fc-b158-48ef-9310-ca2a5ed6a2d8"
      },
      "source": [
        "from imblearn.over_sampling import SMOTE\n",
        "over = SMOTE()\n",
        "X_fs2, ynew = over.fit_resample(X_fs, y)\n",
        "\n",
        "# split into train test sets\n",
        "from sklearn.model_selection import train_test_split\n",
        "X_train2, X_test2, y_train2, y_test2 = train_test_split(X_fs2, ynew,\n",
        "                                                        test_size=0.3,stratify=ynew)\n",
        "model2.fit(X_train2, y_train2)\n",
        "ypred = model2.predict(X_test2)\n",
        "print(\"Accuracy : \",round(accuracy_score(y_test2, ypred)*100,2))\n",
        "print(\" Train Accuracy :\",round(accuracy_score(y_train2, model2.predict(X_train2))*100,2))\n",
        "\n",
        "print(classification_report(y_test2, ypred))"
      ],
      "id": "GjAC0CFa9Y7m",
      "execution_count": 136,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy :  73.44\n",
            " Train Accuracy : 73.83\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.74      0.73      0.73     12394\n",
            "           1       0.73      0.74      0.74     12395\n",
            "\n",
            "    accuracy                           0.73     24789\n",
            "   macro avg       0.73      0.73      0.73     24789\n",
            "weighted avg       0.73      0.73      0.73     24789\n",
            "\n",
            "Accuracy :  72.53\n",
            " Train Accuracy : 73.1\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.86      0.53      0.66     12394\n",
            "           1       0.66      0.92      0.77     12395\n",
            "\n",
            "    accuracy                           0.73     24789\n",
            "   macro avg       0.76      0.73      0.71     24789\n",
            "weighted avg       0.76      0.73      0.71     24789\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "L2ly554tgnHF",
        "outputId": "0b47941e-e930-4475-edb8-015ce65a6989"
      },
      "source": [
        "##checking 0 and 1 observation or verifying balance dataset or not\n",
        "count_arr = np.bincount(y_train2)\n",
        "print('Total occurences of \"0\" in y_train2: ', count_arr[0])\n",
        "print('Total occurences of \"1\" in y_train2: ', count_arr[1])\n",
        "print('ratio zero to one occurence=',int(count_arr[0]/count_arr[1]) )\n",
        "print()\n",
        "count_arr1 = np.bincount(y_test2)\n",
        "print('Total occurences of \"0\" in y_test2: ', count_arr1[0])\n",
        "print('Total occurences of \"1\" in y_test2: ', count_arr1[1])\n",
        "print('ratio zero to one occurence=',count_arr1[0]/count_arr1[1])\n",
        "\n",
        "print('difference=',np.abs((count_arr[0]/count_arr[1])-(count_arr1[0]/count_arr1[1])))\n"
      ],
      "id": "L2ly554tgnHF",
      "execution_count": 137,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Total occurences of \"0\" in y_train2:  28920\n",
            "Total occurences of \"1\" in y_train2:  28919\n",
            "ratio zero to one occurence= 1\n",
            "\n",
            "Total occurences of \"0\" in y_test2:  12394\n",
            "Total occurences of \"1\" in y_test2:  12395\n",
            "ratio zero to one occurence= 0.999919322307382\n",
            "difference= 0.00011525703491888084\n",
            "Total occurences of \"0\" in y_train2:  28920\n",
            "Total occurences of \"1\" in y_train2:  28919\n",
            "ratio zero to one occurence= 1\n",
            "\n",
            "Total occurences of \"0\" in y_test2:  12394\n",
            "Total occurences of \"1\" in y_test2:  12395\n",
            "ratio zero to one occurence= 0.999919322307382\n",
            "difference= 0.00011525703491888084\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5MzANuv0YHhp"
      },
      "source": [
        "\n",
        "\n",
        "```\n",
        "## hyperparamter tuning \n",
        "mlp_gs = MLPClassifier(max_iter=100)\n",
        "parameter_space = {\n",
        "    'hidden_layer_sizes': [(10,30,10),(20,)],\n",
        "    'activation': ['tanh', 'relu'],\n",
        "    'solver': ['sgd', 'adam'],\n",
        "    'alpha': [0.0001, 0.05],\n",
        "    'learning_rate': ['constant','adaptive'],\n",
        "}\n",
        "from sklearn.model_selection import GridSearchCV\n",
        "clf = GridSearchCV(mlp_gs, parameter_space, n_jobs=-1, cv=5)\n",
        "clf.fit(X_train2, y_train2)\n",
        "print('best paras',clf.best_params_)\n",
        "```\n",
        "\n",
        "\n",
        "***MLP Classifier::best paras --> {'activation': 'tanh', 'alpha': 0.05, 'hidden_layer_sizes': (10, 30, 10), 'learning_rate': 'adaptive', 'solver': 'adam'}***"
      ],
      "id": "5MzANuv0YHhp"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wdjmHvKZjztT"
      },
      "source": [
        "\n",
        "**### optimization of MLP Classifier**\n",
        "\n",
        "---\n",
        "\n"
      ],
      "id": "wdjmHvKZjztT"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6XYtC6rZzOeQ"
      },
      "source": [
        "\n",
        "\n",
        "```\n",
        "mlp_gs = MLPClassifier(max_iter=100)\n",
        "parameter_space = {\n",
        "    'hidden_layer_sizes':np.arange(10,20),\n",
        "    'activation': ['tanh', 'relu'],\n",
        "    'solver': ['sgd', 'adam','lbfgs'],\n",
        "    'alpha': [0.0001, 0.05],\n",
        "    'learning_rate': ['constant','adaptive'],\n",
        "}\n",
        "from sklearn.model_selection import GridSearchCV\n",
        "clf = GridSearchCV(mlp_gs, parameter_space, n_jobs=-1, cv=5)\n",
        "clf.fit(X_train2, y_train2)\n",
        "print('best paras',clf.best_params_)\n",
        "```\n",
        "\n"
      ],
      "id": "6XYtC6rZzOeQ"
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YqGD7Vi5Ba9b",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "52382be5-0d27-4b00-d937-8f9c7fc405fb"
      },
      "source": [
        "from sklearn.neural_network import MLPClassifier\n",
        "from sklearn.metrics  import classification_report\n",
        "model2_new = MLPClassifier(activation='tanh',hidden_layer_sizes=(10,30,10),\n",
        "                           learning_rate='adaptive',solver='adam')\n",
        "\n",
        "model2_new.fit(X_train2, y_train2)\n",
        "\n",
        "ypred2 = model2_new .predict(X_test2)\n",
        "\n",
        "from sklearn.metrics import accuracy_score, classification_report\n",
        "\n",
        "print(\"Accuracy : \",round(accuracy_score(y_test2, ypred2)*100,2))\n",
        "print(classification_report(y_test2, ypred2))"
      ],
      "id": "YqGD7Vi5Ba9b",
      "execution_count": 138,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy :  72.48\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.71      0.76      0.74     12394\n",
            "           1       0.74      0.69      0.71     12395\n",
            "\n",
            "    accuracy                           0.72     24789\n",
            "   macro avg       0.73      0.72      0.72     24789\n",
            "weighted avg       0.73      0.72      0.72     24789\n",
            "\n",
            "Accuracy :  74.57\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.80      0.65      0.72     12394\n",
            "           1       0.71      0.84      0.77     12395\n",
            "\n",
            "    accuracy                           0.75     24789\n",
            "   macro avg       0.75      0.75      0.74     24789\n",
            "weighted avg       0.75      0.75      0.74     24789\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gcjBc_4siYEL",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 813
        },
        "outputId": "a5eea25d-e953-4319-d4c5-7ac6b51b497f"
      },
      "source": [
        "print(\"Train Accuracy:\",round(accuracy_score(y_train2,model2_new.predict(X_train2))*100,2))\n",
        "\n",
        "from sklearn import metrics\n",
        "cm = metrics.confusion_matrix(y_test2, ypred2)\n",
        "print(cm)\n",
        "\n",
        "plt.figure(figsize=(6,6))\n",
        "sns.heatmap((cm/np.sum(cm)*100), annot=True, fmt=\".3f\", linewidths=.5,\n",
        "            square = True, cmap = 'viridis');\n",
        "plt.ylabel('Actual label');\n",
        "plt.xlabel('Predicted label');\n",
        "all_sample_title ='Accuracy Score:{0}'.format(accuracy_score(y_test2, ypred2)*100)\n",
        "plt.title(all_sample_title, size = 15);"
      ],
      "id": "gcjBc_4siYEL",
      "execution_count": 139,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Accuracy: 72.77\n",
            "[[9475 2919]\n",
            " [3902 8493]]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAFbCAYAAAAurs6zAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3dd1hT59sH8G/CFFniYLgXigMcKA5cOHBWUSkWK846sW7rxu0Ptah10Wqto1Zrq2IF3FqtVq1aZ92KiykIsldy3j94TY3hQJARkn4/15XrIs855zn3ScidO895ciIRBEEAERHpFKmmAyAioqLH5E5EpIOY3ImIdBCTOxGRDmJyJyLSQfqaDoCIqLSSR9kXug+pzcMiiOQj9quRvRIRUbFi5U5EJEIOeaH70FQFzeRORCRCJhQ+uWsqyXJYhohIB7FyJyISIYf2Xp2FyZ2ISERRjLlrCpM7EZEImRZfV5HJnYhIhDYPy/CEKhGRDmLlTkQkQqbFlTuTOxGRCG0eltGq5O7m5obw8HAcP34c1atX13Q4pcrDhw+xdu1a3Lp1C4mJiahQoQKcnJwwbtw42NsX/voYxeXy5cvw8fHJdZmrqyu+//57AMCRI0dw6NAh/PPPP0hOTkbNmjUxYsQI9O7du0D7mzBhAk6ePIn58+fj888/V7QLgoDdu3djz549ePXqFaysrODm5oZJkybB3NwcAPD69WssWLAA9+7dQ1xcHCwsLNCsWTNMnToVNWrUUPT17v80N3/88QcqVaqkuH/lyhWsW7cOd+7cgb6+PurXr4+VK1fCzs4OALB3714cO3YMDx48QEZGBurWrQtfX1+4uroq9ZuYmIgVK1bg1KlTyMrKQvPmzTF//nyl18mBAwcwe/ZslZgWLlyIzz77rEB9HT16FNu3b0dYWBhSU1NhZ2eHvn37YtSoUTA0NASg/nM7ZMgQ/PXXX7mut3fvXjRt2hQAsHv3bpw9exY3b95EQkICdu7cCRcXF6X11T1GdfGEagm4fv264gUTHByMCRMmaDii0uP58+fw8vJC48aNMX/+fJibm+P58+c4evQoHjx4UKqTe8OGDfHzzz8rtUVERGDKlClo166dom379u2oUqUKZs+ejXLlyuHcuXOYNm0a4uPjMWTIELX2df78edy4cSPXZbt27cLy5csxfvx4tGzZEs+ePcOaNWsQERGBzZs3AwDS09NhYWGBSZMmwc7ODrGxsQgMDMTQoUNx+PBhxZvAhg0bkJmZqdT//PnzIZVKlRL7uXPnMH78eAwaNAgTJkxAeno6rl27hoyMDMU6gYGBaNeuHQYPHowyZcrgt99+w6hRo7Bx40Z07txZsd6UKVPw8OFDzJkzB2ZmZti8eTOGDRuGw4cPw9TUVCmWHTt2wNjYWHG/atWqSsvV6SshIQEuLi4YOXIkzMzMcOvWLWzYsAGxsbFYsGABAPWfWz8/PyQnJyut98033+Du3bto3Lixou3QoUOQSCRwdXVFcHBwrs+jusf4nyBoiSVLlghNmjQRPD09hR49emg6HIXs7GwhIyNDozEEBAQILVu2zDUOuVxe7PtPS0sr0v62bNki1K9fX4iKilK0xcXFqaw3depUoVOnTmr1mZmZKXTv3l3Yt2+fYG9vL+zatUtpuaenp+Dr66vUtmPHDqF+/fpCSkqKaL9hYWGCvb29cOzYMdF1YmJiBAcHB+Hbb79Viqd9+/ZCQEBAnnHndtxeXl7C559/rrj/999/C/b29sKff/6paHv9+rXg6OgobN26VdG2f/9+wd7eXkhOThbdn7p95SYgIEBo3rx5nv9zuT23H8rIyBBatGghLFiwQKldJpMJgiAIDx48EOzt7YVLly6pbKvOMRZE+CvbQt80RStmy8hkMhw5cgRubm4YMGAAnjx5gvv376usd+XKFQwZMgRNmzZF8+bNMWTIENy9e1exPDw8HFOnToWLiwucnJzQp08fHD58GEDOR8h69erh4UPly3MOGTIEX375peL+rFmz0L9/f5w8eRK9evWCo6Mjbt26hZiYGMyePRudO3eGo6Mj3N3dsWbNGpUKLj09HStXrkSnTp3QqFEjuLm54euvvwYArFy5Ep07d4bwwUfBAwcOoFGjRnjz5k2uj09iYiLMzMwUH4ffJ5FIlO6fOHECAwcOhKOjI1xcXPDFF18oDSFcvHgRnp6eaNy4Mdq0aYOFCxciJSVFsfzd4/THH39g7NixaNq0KRYvXgzg36qsZcuWcHJywsiRI/H06dNcY85LSEgIWrRoAWtra0WblZWVynoODg6IiYlRq8+dO3fC2NgYAwYMyHV5dna2SoVrbm4OQRBUno/3WVpaAgCysrJE1zly5Ajkcjl69eqlaPvzzz8RFRUFb2/vPONW57jv3bsHAwMDtGzZUtFWoUIF1K9fH7///nue/X+oMH1ZWlrm+TgAuT+3H/rjjz/w9u1blSE3qbTk05UMQqFvmqIVyf3y5cuIjY1Fz5494e7uDgMDA5WPZZcvX8awYcNgYGCA//3vf1izZg2aN2+O6OhoAEBcXBy8vLxw+/ZtfPXVVwgMDMTAgQMRGRlZ4HjCw8OxatUqjB49Glu2bEGVKlUQHx8PS0tLzJ49G1u3bsXIkSNx4MABLF26VLGdIAgYP3489uzZA29vb2zZsgVffvkl4uPjAQADBgzAq1evVMYfDxw4gE6dOuX6QgeABg0a4OXLl1i6dCkeP34sGndQUBB8fX1RrVo1rF27FsuXL0eNGjUUbxqPHj3CF198gXLlymH9+vWYOHEigoODld7c3pk7dy7q16+PTZs2YeDAgUhISIC3tzfCwsKwcOFCrF27FqmpqRg+fDjS09MV2w0ZMiTPYZSwsDDcvXtXrbH0GzduoGbNmvmu9/r1a2zatAlz5swRTRADBw7E0aNHcfbsWSQnJ+Pu3bv47rvv4OHhgbJlyyqtK5fLkZWVhfDwcCxbtgyVK1dGhw4dRPcfGhqKJk2aoHLlyoq2mzdvwtLSEjdv3kS3bt3QoEED9O7dG6dPny7wcWdmZkIqlUJPT09pPQMDg1zfXLt27YoGDRrA3d0de/fuVVpW0L5kMhnS0tJw9epV7Nq1C5999plKQfGOus9taGgobGxs4OzsnOd6ecnrGAtCJhT+pilaMeYeHBwMc3NztGvXDoaGhmjbti1CQ0Mxbdo0xT9SQEAA6tWrh++//17R1r59e0Uf27dvR3JyMg4cOKAY92zduvVHxZOQkIDt27fDwcFB0WZjY4OvvvpKcb9Zs2YoU6YM5syZg3nz5sHQ0BDnz5/HhQsXsGnTJqXx0n79+gEAateujWbNmuHAgQOKE0UvX77E1atXFeO+ufHw8MCFCxewa9cu7Nq1C5aWlmjfvj18fHwUY5ZyuRxff/01unbtioCAAMW278exadMm2NnZYfPmzYoXt4WFBaZMmYLr168rTmwBQPfu3TF58mTF/bVr1yItLQ1BQUGKarZZs2Zwc3PD/v37MXjwYAD5V18hISEwMDBAt27d8lzv4sWLOHnyJJYvX57negCwatUquLq6okWLFqLreHt7IyUlBWPHjoVcnvOV8y5duig+lbxv4cKFirHkqlWrYtu2bSpV/zvh4eG4ceMG5s6dq9QeGxuLtLQ0zJ8/H1OmTEG1atXw888/w9fXFwcPHkS9evVy7e/XX3/F3bt3lf7XqlWrhoyMDDx48ECxXXp6Oh49eqT0qatixYqYNGkSHB0dIZPJEBoaCj8/P6Snp2PYsGEF6uudJk2aKD6d9uvXDzNnzsw1bkC95zYtLQ2nT5+Gl5eX6JtEXtQ5xoLQ3osPaEHlnpmZiRMnTqBLly6KYYeePXsiPDwc169fBwCkpqbi5s2b8PDwEP2HuHTpEtq1a6d0QutjWVtbKyV2IKcq3759O3r27AlHR0c0bNgQ06dPR2ZmpuLTwaVLl2BpaamUUD80cOBAHD9+XPFCOnDgACpUqKB0AupD+vr6WLt2LQ4dOoRJkyahYcOGOHLkCAYNGqT4KB0WFoaYmBj0799ftJ9bt26hS5cuSlWbu7s79PX1ce3aNaV1O3bsqHT/4sWLaNOmDUxNTZGdnY3s7GyULVsWDRs2xJ07dxTr7dixAzt27BCNITQ0FG3btlW8QeTm1atXmDZtGjp37pzn8QA5J+KPHTuWZ9IBcgqITZs2YfLkyfjxxx+xfPly3L59WyUpA8DYsWPxyy+/YN26dbCyssKIESMQGxuba78hISGQSqXo0aOHUrsgCMjIyMDkyZMxaNAgtGnTBgEBAahSpQq2bt2aa1937tzB0qVL4ePjg1atWinaXV1dUaVKFfj5+eHp06eIiYmBn58fkpKSlF4P7dq1w/jx4+Hq6ooOHTrA398fPXr0wObNmxVvaOr29c7evXuxe/duzJo1C6dOncr1zfAddZ7bM2fOIDU1VWkIqyDUOcb/ilKf3M+dO4fExER06NABiYmJSExMhIuLCwwNDRESEgIgZ8xZEARUrFhRtJ+EhIQ8lxdEhQoVVNp27NiBlStXomvXrti0aRN++eUXxayBd7Mf1ImhR48ekEgkOHLkCARBQFBQEPr27Qt9/fw/ZNWvXx/jx4/Htm3bcPToUVSsWBFr164FAMXQT177f/36tcqx6enpwdLSEm/fvlVqL1++vNL9+Ph4hIaGomHDhkq3y5cvqz30df/+fTx58iTPF3ZCQgK++OIL2NnZYfXq1fn2uXz5cnh5ecHMzEzx/wPkVKNJSUkAcj7VLFmyBD4+PhgzZgxatGiBAQMGYNmyZYrpl++zs7ODo6Mjunfvjm3btiEpKQm7d+/Odf+hoaFwcXFReVzfzax5fyqfnp4eWrRogSdPnqj08/LlS4wZMwatWrXCrFmzlJYZGhpizZo1iI2NRY8ePdCuXTu8fPkSffv2zff/zd3dHQkJCYrzLgXtq2HDhnB2dsbw4cMxb9487NmzBy9evFBZT53nFsh5M6xevbrSLJnC+vAYC0IGSaFvmlLqh2XeJfBJkyapLDt69CjmzJkDc3NzSKVSvH79WrQfS0vLPJcbGRkBUD0x9vbtW5QrVy7fOI8ePQp3d3dMmTJF0fbhizS/GADAxMQEvXr1wsGDB1G5cmVERETkW53mpkqVKujevTt++uknAFAcQ177r1ixIuLi4pTaZDIZEhISYGFhodT+YRVnYWEBNzc3jB8/XqXfD8esxYSEhMDY2Fj0k01aWhrGjh2LrKwsfPvttyhTpky+fYaFheHWrVsqnxZWrVqFgIAA3L17F/Hx8UhISED9+vWV1mnQoAEA4MWLF2jYsGGu/ZuamqJq1ap4+fKlyrKnT5/i3r17WLZsmcqy2rVrA4DKyVpBEFQe27i4OIwcORJ2dnZYs2aNyng4ADg6OuLEiRMICwuDvr4+qlWrhjFjxsDJySnXuN/JrRr/2L7ePV6vXr1CtWrVlJbl99wCQFJSEs6dO4dRo0bluZ+C+pjhnXfk2jvNvXRX7qmpqThz5gx69+6NnTt3Kt1mz56N2NhYXLp0CSYmJnByckJQUJDozIbWrVvj/Pnzoh+fbWxsACgn5MjISLVne6Snp6vMVnk3E+f9GBISEnDmzJk8+xo4cCCuXr2K9evXo0mTJopEIObDhPzOs2fPFBVjzZo1YW1tjaCgINF+nJyccPLkSchkMkXb8ePHkZ2djebNm+cZQ+vWrfH48WPUrVsXjRs3VrrVqlUrz23fCQkJQadOnXJ9M8jOzsakSZPw7NkzbN26VeWTg5jAwECV/x0g58Tu9u3bAeTMSClTpozSzCoAiuGk90+EfujNmzcICwtDlSpVcj0esTFmV1dX6Ovr49KlS4o2mUyGK1euKL3JpKSk4IsvvgCAfN/QJBIJatWqhWrVquHZs2f4888/MXDgQNH1AeDYsWMoV66cyjF+TF9///03AIg+FmLP7TsnTpxAZmZmgb+Ylh+xY1QHK/dicurUKaSlpcHHx0elamjWrBk2b96M4OBgtG3bFtOmTcPw4cMxatQoeHl5oUyZMrhx4wYaNWqETp06YdiwYQgKCsLgwYMxduxY2NjY4OnTp0hNTcUXX3wBGxsbNGrUCOvWrUOZMmUgl8vx7bff5jk++L42bdpg165dcHR0RLVq1XD48GE8f/5caZ22bdvC1dUV06ZNw4QJE9CgQQO8fv0aV69eVRqrdHJyQt26dXHt2rU8xzDf2bRpE+7fv4/evXujdu3aSEtLw/Hjx3HmzBnFiTepVIoZM2Zg+vTpmDZtGnr37g2JRIJLly6hV69eaNy4McaNGwcPDw9MmDABn332GaKiorB69Wq4uroqnUzNzbBhw/Dbb79h6NCh+Pzzz2FtbY3Y2FhcuXIFzZs3V7xghw4dCgAqlfSNGzcQHh6OOXPm5Nr/okWLcPbsWcydOxcJCQlKX0Zq0KCB4o31w/7FZlzUqFFDMd1PIpHg008/xfbt22FsbIxmzZrh5cuX+Oabb9CkSRM0atQIALBt2za8evUKzs7OKF++PF69eoXt27fD0NAQgwYNUtlHaGgo2rdvrxiCeV+lSpXg7e2tmAZbrVo17Nu3D1FRURgzZoxivYkTJ+Lhw4dYsWIFXrx4oTTk0aRJE8XfGzduRK1atVCuXDk8fPgQmzZtQq9evdC2bVulvho3box69epBLpcjNDQUoaGhmDdvntKJbnX6GjlyJNq0aYM6depAT08Pf//9N3744Qf07NlTpWrP77l9JyQkBPXr1xctZm7fvo3w8HBERUUByJn6HB8fj8qVKyuGcdQ9RnVpMjkXVqlO7iEhIahRo0auHwcNDAzQo0cPBAcHY9GiRWjRogW2bduGdevWYcaMGTAwMICDgwO6dOkCIKc627NnD1atWoXly5cjMzMT1atXV3ohBQQEYN68eZgxYwasra0xY8aMPE/+vW/ChAmIj4/HunXrAORMxZo3bx7Gjh2rWEcikWDjxo1Yt24dduzYgTdv3qBSpUro06ePSn+dO3fGy5cv1Tqx1KdPH6SmpuKHH35AdHQ0jI2NUbNmTQQEBCht36dPHxgZGSEwMBBffvml4hPPuymWdevWxZYtWxAQEABfX1+YmpqiV69emDFjRr4xWFlZ4eeff8batWuxYsUKJCYmolKlSmjWrJnSzA+xk1ohISEwMzNTmuH0vgsXLgBArkMcp06dUlSLH3vSbPr06ShXrhwOHTqE7777DlZWVujUqRMmT56sSAr169fH2bNnERoaipSUFNjY2KBly5aYMGGC4pPfO/fu3cPTp0/h6+srus+ZM2fCxMQEmzdvRkJCAho0aIDvv/9eKTm+O+7p06erbP/gwQPF3wkJCVi+fDni4+Nha2uLESNGYMSIEUrr16xZE/v370dUVBQEQUCdOnXg7++vmK1VkL4aN26MgwcPIjw8HHp6eqhatSqmTp2a65tcfs8tkPMJ6NKlS7kOv76ze/duHDx4UHF//fr1AHJmi/3vf/8r0DH+F0iEvL6hQRozcOBA1KxZE6tWrdJ0KET/WTdfFP6yBU7VVM/HlIRSXbn/F92+fRuXLl3C7du3FbNtiEgzOCxDRWbgwIEwNzfHtGnT4OjoqOlwiP7TZKV7zkmemNxLmffHUYmIPhaTOxGRCLnAYRkiIp3DMfcSJI8qvT88QZohtXmIrlJPTYdBpcwJ+S+F7kOmHVdFz5XWJXciopIi1+ITqtobORERiWLlTkQkgmPuREQ6iGPuREQ6SM7KnYhI9/AbqkREVGjjx4/Hq1evIJVKYWJigvnz58PBwQFubm4wNDRU/KjQ9OnT8/zpTYDJnYhIVEmPufv7+8PMzAwAcPLkScyZM0dxmeNvvvkG9vbqf8+HyZ2ISERRzHN//7d732dubq7yQy7vEjsAJCcnF+onApnciYhEyIrg2jI7duzAhg0bVNp9fX0xceJElfa5c+fiwoULEAQBW7duVbRPnz4dgiCgefPmmDp1aq6/8PU+rfuxDl5+gD7Eyw9Qbori8gMHnuT985Lq6FLxrNqV+/uCgoIQEhKCLVu2IDIyEra2tsjMzMSyZcuQkpKC1atX57lfVu5ERCKKYrZMfklcTL9+/bBgwQLFzx0CgKGhIby9vTFu3Lh8t9feeT5ERMVMLkgLfVNXSkoKIiMjFfdPnz4NCwsLGBkZISkpCQAgCAJCQ0Ph4OCQb3+s3ImIRJTkPPe0tDRMmjQJaWlpkEqlsLCwQGBgIOLi4jBx4kTIZDLI5XLUrl0bfn5++fbH5E5EJKIoTqiqq0KFCti3b1+uy4KCggrcH4dliIh0ECt3IiIR2nw9dyZ3IiIRvCokEZEO4lUhiYh0kDZX7tobORERiWLlTkQkgtdzJyLSQfISnOde1JjciYhEaHPlrr2RExGRKFbuREQiCnLhr9KGyZ2ISISM89yJiHQPK3ciIh2kzZW79r4tERGRKFbuREQiOCxDRKSDtPnaMkzuREQitPmqkNr7tkRERKJYuRMRieCwDBGRDuKFw4iIdJA2XziMyZ2ISIQ2V+7a+7ZERESiWLkTEYmQa3H9y+RORCRCpsXDMkzuREQitHnMncmdiEiENl9bRnsjJyIiUazciYhEaPP13JnciYhEcMydiEgHccydiIhKFVbuREQitPl67kzuREQi+CUmIiIdpM1j7kzuREQitHm2jPa+LRERkShW7kREInhClYhIB2nzsAyTOxGRCJ5QJSLSQdpcuWvv2xIREYli5U5EJIInVImIdJA2D8swuRMRiWByJ7XNXApcvAakpQMVrICRnwGevXOWHTkNbPgBiHoN2FYCJn8BdGmXez8rNwGnzgOxbwDrisDowUC/7jnLwl4CqzcD1+8AcjnQqD4w90ugZrWc5SGncvYT+wYwNADauQDzJgGmZYv/+ElV3wnd0W1oR9RoXA2/77mAVSM2AgD0DfQxe/ck2DvXgk2NSpjWyQ+3zt4V7ee3xF1K9w3LGOLw5mPY+OU2uHm7YnLgGMUyiVQCYxMjjHf+Co/+foohfp7wntMfWRnZinVGO01DVFhMER8tlRQm9xL2xWBg6UzA0BB4+hwYOhloUDcn0X+1DNiwLCfZnr0ETPEDTv4MlC+n2k8ZY2DzCqBGVeD2fWD0DKB6FaBpIyApGejUFlg2CyhrAmzaAUyYC4T+/2u/WWPgpw1AOUsgJRVY+DWwbiswd1LJPhaUIy7iDXYv2w9n9yYwMjZUWnbnwj0cWBeC+fum5tvPJ+ZDFH8blzXGvsgtOPfLRQDA6Z/O4/RP5xXLuw3tiMHzBuDR308Vbb///Cf8fdYX9nB0Cit3Ulvdmv/+Lfn//5sX4UC2DDAzBdq3ymnr2Dongb8Izz25Txzx799ODYDmjsCNf3KSu6NDzu2doZ5A4E4J4t8KKGeR86ngfVIp8Dy8aI6PCu78wb8AAPbOtVGxcnlFe3ZWNg6uCwUAyGXyAvXZboALEmLe4vYf93Jd3tWnA07sOvuREf93MLmrIT4+HlFRUQAAGxsblCuXS8b6j1gUAAQdBdIzJHCoK6B9K8DYCKhVHTh9AejQCjjzZ051X692/v2lZ+RU75/1y3351ZtABaucxP7OtVvA2FlAcooEZYwFrF9aNMdGpUNXn46iybtStQpo3L4Bvh65Wam9dR9n7I/9AW8i43Fo41EEBx4viVBLNc6WycOLFy8wf/583L17F5Uq5ZSMMTExaNCgARYtWoQaNWoUdwiljt/UnDHuG/8I+OtGThLX0wP6ugMzlgAZmYCBPrBmEWBSJv/+Fn4N1K8DuLZUXRYVAyxZC8yaoNze3BG4EgpEvxbwSzBgZ1M0x0aaV6laBTh2aICAUZtzXd7VpwPu/HEPUc/+HU8/u+8iQr87ifjot6jvUgcLfp2OlIQUnNl7oaTCLpW0uXIv9i8xzZw5EwMGDMDly5cREhKCkJAQXL58Gf3798dXX31V3LsvtfT0chJs9GtgbxDw51VgdSCwYy1w6ySwcx0wfyVw71He/azaDDwKA9Ys/HeY5503CcCo6TkVfa8uuW9vXTHnTWHaoiI5LCoFugxpj3/O31dK3u/rOqQDTuz8Xantxb1XiIuMh1wux92LD3Hwm1C0G9CqBKKl4lLsyT0hIQGffPIJpNJ/dyWVStG3b1+8ffu2uHdf6mXLgJcRwP3HgLNjzswWqRRo7JAzln7xmvi267cB5y4DW1erznR5mwSMmpZzYnXskNy3f0f2/zGQbug6pAOOf5C832nYph6s7Mrh3K+X8u5EECD5sFr4D5ILkkLfNKXYk7ulpSWCg4MhCIKiTRAE/PbbbzA3Ny/u3ZcqcfE50xBTUnMS6vm/gNBTQKvmOUn92u1/K/W7D4Grt8TH3L/7EQg+CWz7Gkpj6QCQnAJ8MR1o2hiYNkZ128MngIjonL/Do4C1W4FWzYruOKlgpHpSGBgZQKonVfobAAwM9WFgZKDyt5gGre1RvrKVYpbMh7oO7Yjz+y8jLTldqb31J84wtcypEOq1qIN+E3viz9+uFPbQtF5JJ/fx48fjk08+Qb9+/eDt7Y1793JOiIeFhcHLywvu7u7w8vLCs2fP8u1LIryfdYvBs2fP4Ofnh3v37sHa2hoAEB0djfr162PhwoWoVatWgfqTR9kXR5gl4k0CMGkB8OBJzvxzO2vg8wHAp31ylu8+AOz8NWf+uZUl4O0BDPfKWXb4BPDtj0Dwjpz7Dh0kMDAQYPDeWZPRg4ExQ3JO1s5ekXOi9P3i6/COnH2u3QIEHQMSkwBzM6C9CzBltOqbhLaQ2jxEV6mnpsP4aEP8POHj96lS285F+7Br0S/Y9XQjbGooT2/6vOZ4RD9/jc9me6CRqwPm9lquWDYpcDSMyxjBf6jqlEYDIwPsi9yCxQNX4/rpO0rL5uyehObdnGBgZIDXr+JwePMxBK0/UoRHWfJOyH8pdB+dTk8rdB9n3L5We92kpCSYmZkBAE6ePImNGzfi4MGD8PHxwYABA9C3b18cOnQI+/fvx86dO/Psq9iT+ztv3rxBZGQkAMDW1hZWVlYf1Y82J3cqHtqe3Kl4FEVy73hqeqH7+K3FAiQmJqq0m5ub5zl6ERQUhJ07d2LLli1wd3fH5cuXoaenB5lMBhcXFxw/fjzPPFpiUyGtrKw+OqETEWmrHQAomdsAABz9SURBVDt2YMOGDSrtvr6+mDhxokr73LlzceHCBQiCgK1btyIyMhLW1tbQ09MDAOjp6aFSpUqIjIwsHcmdiEjbFMU896FDh8LDw0OlXaxqX7ZsGYCcyn3lypWYNOnjvjrO5E5EJKIoZrvkN/wipl+/fliwYAFsbGwQHR0NmUymGJaJiYmBra1tntvzxzqIiEQIgqTQN3WlpKQozksCwOnTp2FhYYHy5cvDwcEBwcHBAIDg4GA4ODjkO8zNyp2IqBRIS0vDpEmTkJaWBqlUCgsLCwQGBkIikWDhwoWYNWsWNm3aBHNzc/j7++fbH5M7EZGIkvwSUoUKFbBv375cl9WuXRu//FKw2T9M7kREIgoyrFLaMLkTEYnQ5guHMbkTEYkoma94Fg/OliEi0kGs3ImIRPDHOoiIdBBPqBIR6SCeUCUi0kE8oUpERKUKK3ciIhEccyci0kFM7kREOkibT6hyzJ2ISAexciciEqHNs2WY3ImIRHDMnYhIBzG5ExHpIC0eleEJVSIiXcTKnYhIBIdliIh0kRaPyzC5ExGJYOVORKSDtHmeO0+oEhHpINHK3dvbGxJJ/h9Jdu/eXaQBERGVFjo5LOPp6VmScRARlT66mNw9PDxKMg4iolJH58fcBUHAvn374OPjgz59+gAArly5gtDQ0GINjoiIPo5ayX3dunX49ddf4eXlhcjISACAjY0Ntm7dWqzBERFplFAENw1RK7kfPHgQgYGB6NWrl+Ika5UqVfDy5ctiDY6ISJMEQVLom6aoNc9dJpOhbNmyAKBI7ikpKTAxMSm+yIiINE3Xx9w7dOiAFStWIDMzE0DOGPy6devQqVOnYg2OiEiTtLlyVyu5z549G69fv0bz5s2RlJSEpk2bIiIiAtOnTy/u+IiI6COoNSxjamqKjRs3Ii4uDuHh4bC1tUXFihWLOzYiIs3S4mEZta8tk5iYiAsXLiAmJgaVKlVChw4dYGFhUZyxERFpmPZ+iUmtYZmLFy/Czc0Nu3btwu3bt/Hjjz+ic+fOuHjxYnHHR0SkOVo8FVKtyn3JkiVYvHgxevbsqWg7cuQIFi1ahKNHjxZbcEREGqXFwzJqVe4xMTFwd3dXauvatStiY2OLJSgiIioctZJ73759Va7+uGfPHvTr169YgiIiKhUESeFvGqLWJX/lcjn27t2LrVu3wtraGtHR0YiLi4OTk1OJBUpEVNK0+cJhal/y99NPPy32YIiIShVdTO685C8RkfZSe557bGwsbt26hfj4eAjvfVYZOHBgsQRGRKRxuvhjHe87efIkZsyYgerVq+Px48eoU6cOHj16hGbNmjG5E5HOkujisMz71q5di+XLl6NHjx5o0aIFgoKCsH//fjx+/Li44yMi0hwtTu5qTYWMiIhAjx49lNo8PDwQFBRULEEREZUKWjwVUq3kXr58ecUXlipXrozr16/jxYsXkMvlxRocERF9HLWGZTw9PXHt2jW4u7tj2LBh8PHxgVQqxfDhw4s7PiIizdHiYRm1kvvo0aMVf/fr1w8tW7ZEWloaateuXWyBERFpnK4n9w/Z2dkVdRxERKWPLib3Dh06KC4/kJfff/+9KOMhIqIiIJrcV61aVZJxEBGVPrr4JaaWLVuWZBxERKWOzn+JiYjoP6kEk3t8fDxmzpyJFy9ewNDQENWrV8fixYthZWWFevXqwd7eHlJpzuz1lStXol69enn2x+RORFQKSCQSjBo1Ci4uLgAAf39/rF69GsuXLwcA7N27F2XLllW7PyZ3IqJilJiYiMTERJV2c3NzmJubK+5bWloqEjsANGnSBHv27Pno/WpdcpfaPNR0CFQKnZD/oukQSAcVxZj7jh07sGHDBpV2X19fTJw4Mddt5HI59uzZAzc3N0XbkCFDIJPJ0L59e0ycOBGGhoZ57lciCLn/1siMGTPUmgq5cuXKfNcpSu4tFpXo/qj0O3bFD1sfttN0GFTKjLL/o9B91FoXUOg+bgwfpVbl/r5FixYhOjoaGzZsgFQqRWRkJGxtbZGcnIwZM2bA3t4eU6ZMyXO/opV79erVC3gIREQ6pggq97ySeG78/f3x/PlzBAYGKk6g2traAgBMTU3h6emJH374Id9+RJO7r6+v2sEQEemkEp4KGRAQgDt37uC7775TDLu8ffsWRkZGMDY2RnZ2No4dOwYHB4d8+1J7zD0zMxNhYWEqv8TUunXrjzgEIiJ636NHj/Dtt9+iRo0aGDRoEACgSpUqGDVqFBYsWACJRILs7Gw0bdoUkyZNyrc/tZL71atXMXnyZGRmZiI5ORmmpqZISUmBjY0NTp06VbgjIiIqpUryS0x169bFgwcPcl12+PDhAven1vXcV6xYgVGjRuGvv/5C2bJl8ddff2HcuHHw9vYu8A6JiLSGUAQ3DVEruT979gw+Pj5KbaNHj8b27duLIyYiotJB15O7mZkZkpOTAQAVK1bE48ePkZiYiNTU1GINjoiIPo5aY+5du3bF2bNn0adPHwwYMAA+Pj7Q19eHu7t7ccdHRKQxOn/hsLlz5yr+HjlyJJycnJCSkoJ27fjFESLSYbp4yd+8ODs7F3UcRESlj65X7t7e3qKXIti9e3eRBkREVFro/LCMp6en0v3Xr19j//796NOnT7EERUREhaNWcvfw8FBpc3d3x+zZs3mZAiLSXbpeuefG2tpa9NtURES6QOeHZX799Vel++np6Th+/DiaNGlSLEEREZUKup7cDx06pHTfxMQETZs2xbBhw4ojJiKi0kHXk/uuXbuKOw4iIipCal1+oGXLlrm283K/RKTLJELhb5qiVnLPysrKtU0ulxd5QEREVHh5Dsu8+/JSZmYmBg8erLQsKioKTZs2LdbgiIg0SlfH3D09PSEIAm7fvo2BAwcq2iUSCcqXL49WrVoVe4BERFRweSb3d19ecnJyQu3atUskICKi0kKb57mrNea+Z88e/P3330ptf//9N5YtW1YsQRERlQq6/mMdwcHBaNSokVJbo0aNEBwcXCxBERGVClqc3NWa5y6RSCAIylHKZDLOliEinabzwzLOzs5Yu3atIpnL5XKsX7+e13UnIiql1P4lpjFjxsDV1RV2dnaIjIxExYoVsXnz5uKOj4hIc7S4clcrudvY2ODgwYO4efMmoqKiYGtrC0dHx+KOjYhIo3R+WAYApFIpmjZtih49eqBMmTJYtWoV2rdvX5yxERFplq6fUAWAN2/e4PDhwwgKCsL9+/fRvHlzpR/OJiLSOVpcueeZ3LOysnD69GkcPHgQ58+fR7Vq1dCrVy9ERERg3bp1KF++fEnFSUREBZBncm/bti0kEgn69++PiRMnomHDhgByvtRERKTrdHbMvV69ekhKSsLNmzdx+/ZtvH37tqTiIiLSPF0dc9+1axfCw8MRFBSEbdu2YenSpXB1dUVqaiqys7NLKkYiIs3Q1codACpXrowJEybg+PHj2L59OypWrAipVIpPPvkEK1euLIkYiYiogNSeLQPkfFPV2dkZ8+bNw4kTJxAUFFRccRERaZw2j7kXKLm/Y2RkhN69e6N3795FHQ8RUenxX0vuRET/Bf+5yp2I6D9Bi5O72pcfICIi7cHKnYhIjBZX7kzuREQiJJoOoBCY3ImIxLByJyLSPdo8W4YnVImIdBArdyIiMVpcuTO5ExGJYXInItI9HHMnIqJShZU7EZEYLa7cmdyJiERo87AMkzsRkRgmdyIi3aPNlTtPqBIR6SBW7kREYrS4cmdyJyISw+RORKR7tHnMncmdiEhMCSb3+Ph4zJw5Ey9evIChoSGqV6+OxYsXw8rKCjdu3MCCBQuQkZGBypUrY9WqVShfvnye/fGEKhFRKSCRSDBq1CgcO3YMhw8fRtWqVbF69WrI5XLMmDEDCxYswLFjx+Ds7IzVq1fn2x+TOxGRCIkgFPqWmJiIV69eqdwSExOV9mVpaQkXFxfF/SZNmiAiIgJ37tyBkZERnJ2dAQCDBg3C0aNH842dwzJERGKKYFhmx44d2LBhg0q7r68vJk6cmOs2crkce/bsgZubGyIjI2FnZ6dYZmVlBblcjoSEBFhaWorul8m9BH3i2QJdezdBjTqV8PvxO/h60SEAgLWtBXb+NhlpqZmKdfftvICfvj+n0kdFa3Ns2TdBqa2MiSG+W3sc+3dfVGqfOv8TuH/SFMM9vkHEq/g8YyDNCv76LV7czEJWuoCy5aRo2d8Eju5llNb5c08KLvyUAs8llqjRxFClj5QEOU5vScLLOzn9VKiuj04jTWFXz0CxTupbOU5/l4QnVzMhkQC1nA3Re7oFAOD+H+m49lsqYp5mw9beAINWlCveg9YCRXFCdejQofDw8FBpNzc3F91myZIlMDExweeff44TJ0581H6Z3EtQXGwSftp2Ds6t6sDQWPWh7+/2P8hlef83vY5ORL8OKxT3re0s8cOBiTh/+q7Seg2dqsK2iuqLM78YSDNaDSyL7l/qQd9AgriX2dg7JwGVauvDpk5OYo6PzMaDC+koayU+kpqVLsCmrgE6jTSFiYUUt0+k48CiBIz+vjwMy+RsF7T8LWzq6mPstvLQN5Ig9nm2YntjMymaf2KCN69keHErU2w3VEDm5uZ5JvIP+fv74/nz5wgMDIRUKoWtrS0iIiIUy9+8eQOpVJpn1Q5wzL1EXThzHxfPPkDi29Qi67NLTyfcuf4c0ZFvFW1SPQnGz+iBTauOlEgMVHgVqutD30ACAJBIcm4JkTLF8pOByegwzBR6ebwfW9rooUU/E5ha6UGqJ4FT9zKQZQNvwnP6Cfs7A0mxMnQcbgqjslLo6UtgXfvfqr5GE0PUb2cM0zzeQP5zhCK4FUBAQADu3LmDjRs3wtAw59NZo0aNkJ6ejqtXrwIA9u7di+7du+fbF0u3UmTXb5MhCMD1v55gy7oTSHyblu82XXo5qgzf9P+sNW5ff4GwxzHFFSoVgxObknDnVBqyM4FKtfRRyznnxf3gfDr0DYBazkYAktTuL/ppFmTZAsrZ6gEAIh9kw6qyPkLXJiLsWiYsrfXQcYQpqjZWHeKhHCU5z/3Ro0f49ttvUaNGDQwaNAgAUKVKFWzcuBErV66En5+f0lTI/DC5lwJvE1Lh6/MdnjyMgrmFCXxn9sRXS/pj7pe789yuUZNqKGdlij9O/TskU9HaHL36N8eEId8Vd9hUxLqON0PnMaaIuJ+Fl3eyoGcgQWaqHOd2puDTJXl/BP9QRqocoQGJaPNZWRiVzanEk+JkeHY9E+4TzdBjkjke/pmBg8veYtS35WFiwWo9VyWY3OvWrYsHDx7kuqxZs2Y4fPhwgfrT6DPap08fTe6+1EhPy8Kje5GQywQkvEnBxlWhcG5dB2VM8q6ouvRywvnTd5GelqVoGzvVHbu3nkVqSkZxh03FQKonQZWGhkiKleFGaBou7ElBw07GsLDWU7uPrAwBBxa/hV09A7TyLKto1zeUwKKSFI7dykBPXwKH9sYwqyBF+L2sPHr7b5MIhb9pSrFX7o8fPxZdFh8fX9y710rC//9DSCQS0XUMjfTRvksDLJrxs1J7kxa10NCpGkZ+2VXRtmbbSAR+fRRnjt0plnip6MllQEKUDC/vZCE5VobroTlDdGmJchz2f4uWA0zgMrCsynbZWQKClr2FWQUpuk0wU1pWsYY+nvz1wZu++L8YabliT+69e/dG5cqVIQiqb2EJCQnFvftSRaongZ6eFFI9KaRSCQwM9SCTyVG3vh1SktMR/iIOpuZlMH56d9y8GpZn9d22Y30kJabj5tVnSu0jBqyHVPrvK3bv0enwm7oHTx9F5xlDfrN0qPikJMjx4lYmarcwhL6hBM9vZuL+uXT0nmGBNoPKQvbec/Pj1Hh0GmmKms1VP9XJsgX8tuIt9A2BnlPMIZEqZ+66rY3w+7Zk3DmVhgYdjfHoUgaSY+Wo7JBzUlUuEyCXAXJ5ToGRnSlAIgX09P/D7wBa/LIo9uReuXJl/PTTT7C2tlZZ1qFDh+LefaniPaI9hozuqLjfpacTdn33O149j8Pw8W6wtCqLlJQMXL/8FCvm7Ves9+WsXgCAb/4X8u+2vZxw6sgtlX28jVedBZOYkIrMjOw8Y/hxy9nCHh59JIkEuBGahhObkiDIAfNKUnT6wgx1XIxU15UCRqZSxdTG4xtzvuXYbYI5Iu5l4cmVTOgbAt8MilVsM3ChBao0NEQZMyn6z7fAic1JOBmYDKsqevCYZ6EYb797Jh1H1v17wnbNgNdo6GaMnlPUn8ana7T5wmESIbeSugj5+/uja9euaNasmcqypUuXYt68eQXqz73FoqIKjXTEsSt+2PqwnabDoFJmlP0fhe6j1eCvC93Hpd3TCt3Hxyj2yv2rr74SXVbQxE5EVJK0uXLn/CciIh3Eee5ERGK0uHJnciciEiGRazqCj8fkTkQkRosrd465ExHpIFbuREQitHm2DJM7EZGY4v0aULFiciciEsHKnYhIF2lxcucJVSIiHcTKnYhIBIdliIh0EU+oEhHpHm2u3DnmTkSkg1i5ExGJ0eLKncmdiEiENg/LMLkTEYmRa292Z3InIhKjvbmdJ1SJiHQRK3ciIhEccyci0kX8EhMRke5h5U5EpIu0OLnzhCoRkQ5i5U5EJELCMXciIh0k13QAH4/JnYhIhDZX7hxzJyLSQazciYjEaG/hzuRORCRKi4dlmNyJiETwS0xERLpIiyt3nlAlItJBrNyJiERIOM+diEgHafGwDJM7EZEY7c3tTO5ERGL4DVUiIipVWLkTEYnR4sqdyZ2ISAxnyxAR6R6OuRMRUanCyp2ISIwWV+5M7kREYpjciYh0EE+oEhHpHp5QJSKiUoWVOxGRmBKs3P39/XHs2DGEh4fj8OHDsLe3BwC4ubnB0NAQRkZGAIDp06ejXbt2+fbH5E5EJKYEk3vnzp3h4+ODwYMHqyz75ptvFMleXUzuRERiiiC5JyYmIjExUaXd3Nwc5ubmivvOzs6F3tf7mNyJiMQUwWyZHTt2YMOGDSrtvr6+mDhxolp9TJ8+HYIgoHnz5pg6darSm4IYJnciomI0dOhQeHh4qLSrk6ABYPfu3bC1tUVmZiaWLVuGxYsXY/Xq1flux+RORCSiKKZCfjj8UlC2trYAAENDQ3h7e2PcuHFqbcfkTkQkRsPz3FNTUyGTyWBmZgZBEBAaGgoHBwe1tmVyJyISIy+55L506VIcP34csbGxGD58OCwtLREYGIiJEydCJpNBLpejdu3a8PPzU6s/JnciolJg3rx5mDdvnkp7UFDQR/XH5E5EJEaLLz/A5E5EJIbJnYhIBzG5ExHpoBI8oVrUJIKgxW9NRETFqEet6YXu48jT/L9wVBxYuRMRiRG099c6mNyJiMRo8cAGkzsRkRgtHnNnciciEqPFlTt/Zo+ISAexciciEqPFlTuTOxGRGCZ3IiIdJNfeqZAccyci0kGs3ImIxGjxsAwrdy0UFhYGLy8vuLu7w8vLC8+ePdN0SKRh/v7+cHNzQ7169fDw4UNNh6M7BKHwNw1hctdCfn5+8Pb2xrFjx+Dt7Y0FCxZoOiTSsM6dO2P37t2oXLmypkPRLXKh8DcNYXLXMnFxcbh79y569+4NAOjduzfu3r2LN2/eaDgy0iRnZ2fFDylT0REEeaFvmsLkrmUiIyNhbW0NPT09AICenh4qVaqEyMhIDUdGRKUJT6gSEYnhtWWopNja2iI6OhoymQx6enqQyWSIiYnhR3Ki4sDZMlRSypcvDwcHBwQHBwMAgoOD4eDgACsrKw1HRqSD5PLC3zSEv8SkhZ48eYJZs2YhMTER5ubm8Pf3R61atTQdFmnQ0qVLcfz4ccTGxqJcuXKwtLRESEiIpsPSet3Nhxe6j6OJPxRBJAXH5E5EJEKbkzvH3ImIRAhafG0ZJnciIjFaPLDB5E5EJEaLp0JytgwRkQ5i5U5EJEaDlw8oLFbuVOJmzZqFNWvWAACuXr0Kd3f3EtlvvXr18Pz581yXDRkyBL/88ota/bi5ueHPP//8qBgKsy2VPEEuFPqmKUzulCs3Nzc4OjqiadOmaNOmDWbNmoWUlJQi34+zszOOHTuW73oHDhzAZ599VuT7J8qTIC/8TUOY3ElUYGAgrl+/joMHD+LOnTvYvHmzyjrZ2dkaiIyoZLByJ51mbW2Ndu3a4dGjRwByhjd2796Nbt26oVu3bgCAM2fOoG/fvnB2dsagQYNw//59xfZ3796Fh4cHmjZtismTJyMjI0Ox7PLly2jfvr3ifmRkJHx9fdGqVSu4uLhg8eLFePLkCfz8/HDjxg00bdoUzs7OAIDMzEz4+/ujY8eOaNOmDRYsWID09HRFX1u3boWrqytcXV3x66+/qn28L168gI+PD1xcXODi4oJp06YhMTFRaZ3bt2+jZ8+eaNGiBWbPnq10THk9FkQlRiDKRadOnYQLFy4IgiAIERERQs+ePYU1a9YIgiAI9vb2wrBhw4T4+HghLS1N+Oeff4RWrVoJN27cELKzs4UDBw4InTp1EjIyMoSMjAyhY8eOwg8//CBkZmYKR44cERo0aCAEBAQIgiAIly5dEtq1aycIgiBkZ2cLffr0EZYtWyakpKQI6enpwpUrVwRBEIT9+/cLgwYNUopx2bJlwpgxY4T4+HghKSlJGDNmjLB69WpBEATh7NmzQuvWrYUHDx4IKSkpwtSpUwV7e3vh2bNnuR7v559/Luzbt08QBEF49uyZcP78eSEjI0OIi4sTvL29haVLlyo9Nr169RIiIiKE+Ph4wcvLS3E8eT0WHz6uRMWJlTuJmjBhApydneHt7Y0WLVpg7NiximWjR4+GpaUljI2N8fPPP8PLywtOTk7Q09ODh4cHDAwMcOPGDdy8eRNZWVkYOnQoDAwM0L17dzRu3DjX/d26dQsxMTGYOXMmTExMYGRkpKjSPyQIAvbt24c5c+bA0tISpqamGDNmjOJ6KkeOHEH//v1hb28PExMT+Pr6qn3c1atXR9u2bWFoaAgrKysMHz4cV65cUVpn8ODBsLW1haWlJcaNG6fYb16PBVFJ4lRIErVx40a0adMm12XvX2I4IiICQUFB+PHHHxVtWVlZiImJgUQigbW1NSQSiWKZnZ1drn1GRkbCzs4O+vr5/1u+efMGaWlp6N+/v6JNEATI///r4jExMWjUqJFiWUF+fi42NhbLli3D1atXkZKSAkEQYG5urrTO+8dvZ2eHmJgYAHk/FkQlicmdPsr7ydrW1hZjx47FuHHjVNb766+/EB0dDUEQFNtERESgatWqKuva2toiMjIS2dnZKgn+/f0BQLly5WBsbIyQkBBYW1ur9PXhr1NFRESofWwBAQGQSCQ4fPgwLC0tcfLkSSxevFhpnQ/7rlSpkuIYxB4LopLEYRkqNE9PT+zduxc3b96EIAhITU3F77//juTkZDRp0gT6+vrYuXMnsrKycPz4cdy+fTvXfhwdHVGxYkV8/fXXSE1NRUZGBq5duwYg5zr20dHRyMzMBABIpVJ4enpi+fLliIuLAwBER0fjjz/+AAB0794dBw8exOPHj5GWloYNGzaofTwpKSkwMTGBmZkZoqOjsXXrVpV1fvrpJ0RFRSEhIQGBgYHo2bNnvo8FUUlicqdCa9y4MZYsWYLFixejRYsW6NatGw4cOAAAMDQ0xPr163Hw4EG0bNkSoaGh6Nq1a6796OnpITAwEM+fP0enTp3Qvn17HDlyBADQqlUr1KlTB66urnBxcQEAzJgxA9WrV8enn36KZs2aYdiwYQgLCwMAdOjQAUOHDsXQoUPRtWtXtGrVSu3j8fX1xd27d+Hs7IzRo0crZgS9r3fv3hgxYgS6dOmCatWqKSr1vB4LopLE67kTEekgVu5ERDqIyZ2ISAcxuRMR6SAmdyIiHcTkTkSkg5jciYh0EJM7EZEOYnInItJB/wd+33H6zYnnAwAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x432 with 2 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Accuracy: 74.8\n",
            "[[ 8089  4305]\n",
            " [ 1998 10397]]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x432 with 2 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 143
        },
        "id": "EGsLr-S-qypG",
        "outputId": "8b6d3424-b0a9-4211-afa0-a05bed082054"
      },
      "source": [
        "from sklearn.metrics import accuracy_score, f1_score,recall_score,precision_score, confusion_matrix\n",
        "acc = accuracy_score(y_test2, ypred2)\n",
        "prec = precision_score(y_test2, ypred2)\n",
        "rec = recall_score(y_test2, ypred2)\n",
        "f1 = f1_score(y_test2, ypred2)\n",
        "results2 = pd.DataFrame([['MLP Classifier', acc*100,prec*100,rec*100,f1*100]],\n",
        "                       columns=['Model', 'Accuracy', 'Precision', 'Recall','F1 Score'])\n",
        "results2"
      ],
      "id": "EGsLr-S-qypG",
      "execution_count": 140,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>Model</th>\n",
              "      <th>Accuracy</th>\n",
              "      <th>Precision</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1 Score</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>MLP Classifier</td>\n",
              "      <td>72.483763</td>\n",
              "      <td>74.421661</td>\n",
              "      <td>68.519564</td>\n",
              "      <td>71.348763</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "            Model   Accuracy  Precision     Recall   F1 Score\n",
              "0  MLP Classifier  72.483763  74.421661  68.519564  71.348763"
            ]
          },
          "metadata": {},
          "execution_count": 67
        },
        {
          "output_type": "execute_result",
          "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>Model</th>\n",
              "      <th>Accuracy</th>\n",
              "      <th>Precision</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1 Score</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>MLP Classifier</td>\n",
              "      <td>74.573399</td>\n",
              "      <td>70.71827</td>\n",
              "      <td>83.880597</td>\n",
              "      <td>76.739122</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "            Model   Accuracy  Precision     Recall   F1 Score\n",
              "0  MLP Classifier  74.573399   70.71827  83.880597  76.739122"
            ]
          },
          "metadata": {},
          "execution_count": 140
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cB3_WaeZLca2"
      },
      "source": [
        "\n",
        "\n",
        "---\n",
        "\n",
        "### `SMOTE:: Model 4`"
      ],
      "id": "cB3_WaeZLca2"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "elFm4QQOcE5M"
      },
      "source": [
        "\n",
        "\n",
        "```\n",
        "## hyperparamter tuning \n",
        "parameter_space = {'bagging_fraction': (0.5, 0.8),\n",
        "            'bagging_frequency': (5, 8),\n",
        "            'feature_fraction': (0.5, 0.8),\n",
        "            'max_depth': (10, 15, 20),\n",
        "            'num_leaves': (1000, 1250, 1500)}\n",
        "from sklearn.model_selection import GridSearchCV\n",
        "clf = GridSearchCV(model4, parameter_space, n_jobs=-1, cv=5)\n",
        "clf.fit(X_train2, y_train2)\n",
        "print('best paras',clf.best_params_)\n",
        "```\n",
        ">>*best paras {'bagging_fraction': 0.5, 'bagging_frequency': 5, 'feature_fraction': 0.8, 'max_depth': 20, 'num_leaves': 1500}*\n"
      ],
      "id": "elFm4QQOcE5M"
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "S9QJejnaaBeV",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "645f1fee-9e8f-4d44-bfcf-62d3d6e87f69"
      },
      "source": [
        "import lightgbm as lgb\n",
        "model4_new = lgb.LGBMClassifier(bagging_fraction= 0.5,\n",
        "                                bagging_frequency= 5, \n",
        "                                feature_fraction= 0.8, max_depth= 20, \n",
        "                                num_leaves= 1500)\n",
        "model4_new .fit(X_train2, y_train2)\n",
        "\n",
        "y_pred2 = model4_new .predict(X_test2)\n",
        "from sklearn.metrics import accuracy_score, classification_report\n",
        "print(\"Accuracy : \",round(accuracy_score(y_test2, y_pred2)*100,2))\n",
        "print(classification_report(y_test2, y_pred2))"
      ],
      "id": "S9QJejnaaBeV",
      "execution_count": 141,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy :  98.52\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.99      0.98      0.99     12394\n",
            "           1       0.98      0.99      0.99     12395\n",
            "\n",
            "    accuracy                           0.99     24789\n",
            "   macro avg       0.99      0.99      0.99     24789\n",
            "weighted avg       0.99      0.99      0.99     24789\n",
            "\n",
            "Accuracy :  98.44\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.99      0.98      0.98     12394\n",
            "           1       0.98      0.99      0.98     12395\n",
            "\n",
            "    accuracy                           0.98     24789\n",
            "   macro avg       0.98      0.98      0.98     24789\n",
            "weighted avg       0.98      0.98      0.98     24789\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AWV6BD5uatlk",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 813
        },
        "outputId": "d795cb4f-2bc4-42a6-f1f2-2bd9efb0d558"
      },
      "source": [
        "print('Train Accuracy:',round(accuracy_score(y_train2,model4_new.predict(X_train2))*100,2))\n",
        "\n",
        "from sklearn import metrics\n",
        "cm = metrics.confusion_matrix(y_test2, y_pred2)\n",
        "print(cm)\n",
        "\n",
        "plt.figure(figsize=(6,6))\n",
        "sns.heatmap((cm/np.sum(cm)*100), annot=True, fmt=\".3f\", linewidths=.5,\n",
        "            square = True, cmap = 'viridis');\n",
        "plt.ylabel('Actual label');\n",
        "plt.xlabel('Predicted label');\n",
        "all_sample_title ='Accuracy Score:{0}'.format(accuracy_score(y_test2, y_pred2)*100)\n",
        "plt.title(all_sample_title, size = 15);"
      ],
      "id": "AWV6BD5uatlk",
      "execution_count": 142,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Accuracy: 100.0\n",
            "[[12151   243]\n",
            " [  124 12271]]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x432 with 2 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Accuracy: 100.0\n",
            "[[12120   274]\n",
            " [  113 12282]]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x432 with 2 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "np5olfzOrEdY"
      },
      "source": [
        "acc = accuracy_score(y_test2, y_pred2)\n",
        "prec = precision_score(y_test2, y_pred2)\n",
        "rec = recall_score(y_test2, y_pred2)\n",
        "f1 = f1_score(y_test2, y_pred2)\n",
        "results2.loc[len(results2.index)] =['LGBM classifier', acc*100,\n",
        "                                  prec*100,rec*100,f1*100]"
      ],
      "id": "np5olfzOrEdY",
      "execution_count": 143,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3pi6IeV7r033"
      },
      "source": [
        "\n",
        "\n",
        "---\n",
        "\n",
        "## `SMOTE: Random Forest`"
      ],
      "id": "3pi6IeV7r033"
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "MSzQBU4ZsVTs",
        "outputId": "4a62a347-a903-48bd-97cd-e8fa1bdcb27d"
      },
      "source": [
        "from sklearn.impute import SimpleImputer\n",
        "\n",
        "imputer = SimpleImputer(strategy=\"mean\")\n",
        "X_imp = imputer.fit_transform(X)\n",
        "X1= pd.DataFrame(X_imp,columns=data_keys[:len(data_keys)-1])\n",
        "\n",
        "X1= X1[['Attr5', 'Attr27', 'Attr29', 'Attr34', 'Attr35', 'Attr40', 'Attr44',\n",
        "       'Attr46', 'Attr47', 'Attr50', 'Attr56', 'Attr58']]\n",
        "\n",
        "\n",
        "from imblearn.over_sampling import SMOTE\n",
        "over = SMOTE()\n",
        "X1, y = over.fit_resample(X1, y)\n",
        "\n",
        "# split into train test sets\n",
        "from sklearn.model_selection import train_test_split\n",
        "X_train2, X_test2, y_train2, y_test2 = train_test_split(X1, y,\n",
        "                                                        test_size=0.3,stratify=y)\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "model_new = RandomForestClassifier()\n",
        "model_new.fit(X_train2, y_train2)\n",
        "\n",
        "y_pred2 = model_new.predict(X_test2)\n",
        "from sklearn.metrics import accuracy_score, classification_report\n",
        "print(\"Accuracy : \",round(accuracy_score(y_test2, y_pred2)*100,2))\n",
        "\n",
        "print(classification_report(y_test2, y_pred2))\n",
        "print('Train Accuracy:',round(accuracy_score(y_train2,model_new.predict(X_train2))*100,2))\n",
        "\n",
        "from sklearn import metrics\n",
        "cm = metrics.confusion_matrix(y_test2, y_pred2)\n",
        "print(cm)\n",
        "\n",
        "plt.figure(figsize=(6,6))\n",
        "sns.heatmap((cm/np.sum(cm)*100), annot=True, fmt=\".3f\", linewidths=.5,\n",
        "            square = True, cmap = 'viridis');\n",
        "plt.ylabel('Actual label');\n",
        "plt.xlabel('Predicted label');\n",
        "all_sample_title ='Accuracy Score:{0}'.format(accuracy_score(y_test2, y_pred2)*100)\n",
        "plt.title(all_sample_title, size = 15);"
      ],
      "id": "MSzQBU4ZsVTs",
      "execution_count": 144,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy :  97.46\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.98      0.97      0.97     12395\n",
            "           1       0.97      0.98      0.97     12394\n",
            "\n",
            "    accuracy                           0.97     24789\n",
            "   macro avg       0.97      0.97      0.97     24789\n",
            "weighted avg       0.97      0.97      0.97     24789\n",
            "\n",
            "Train Accuracy: 100.0\n",
            "[[12053   342]\n",
            " [  288 12106]]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x432 with 2 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy :  97.53\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.98      0.97      0.98     12394\n",
            "           1       0.97      0.98      0.98     12395\n",
            "\n",
            "    accuracy                           0.98     24789\n",
            "   macro avg       0.98      0.98      0.98     24789\n",
            "weighted avg       0.98      0.98      0.98     24789\n",
            "\n",
            "Train Accuracy: 100.0\n",
            "[[12034   360]\n",
            " [  253 12142]]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x432 with 2 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "k5ibmkSUs06S"
      },
      "source": [
        "acc = accuracy_score(y_test2, y_pred2)\n",
        "prec = precision_score(y_test2, y_pred2)\n",
        "rec = recall_score(y_test2, y_pred2)\n",
        "f1 = f1_score(y_test2, y_pred2)\n",
        "results2.loc[len(results2.index)] =['Random Forest Classifier', acc*100,\n",
        "                                  prec*100,rec*100,f1*100]"
      ],
      "id": "k5ibmkSUs06S",
      "execution_count": 145,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-dfRnC6CtbWA"
      },
      "source": [
        "*## Model is choosen for easy deployment on heroku*"
      ],
      "id": "-dfRnC6CtbWA"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8w0Vru9SjvCS"
      },
      "source": [
        "\n",
        "\n",
        "---\n",
        "\n",
        "\n",
        "\n",
        "---\n",
        "\n",
        "\n",
        "\n",
        "---\n",
        "\n"
      ],
      "id": "8w0Vru9SjvCS"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "N8WiPvzAteJf"
      },
      "source": [
        "# Summary of Model: with Oversampling (SMOTE)"
      ],
      "id": "N8WiPvzAteJf"
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 142
        },
        "id": "4VfCCPZ3tmlX",
        "outputId": "528b1954-931e-4ea0-8f61-ca6d16e9e82f"
      },
      "source": [
        "results2.head()"
      ],
      "id": "4VfCCPZ3tmlX",
      "execution_count": 147,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>Model</th>\n",
              "      <th>Accuracy</th>\n",
              "      <th>Precision</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1 Score</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>MLP Classifier</td>\n",
              "      <td>74.573399</td>\n",
              "      <td>70.718270</td>\n",
              "      <td>83.880597</td>\n",
              "      <td>76.739122</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>LGBM classifier</td>\n",
              "      <td>98.438824</td>\n",
              "      <td>97.817776</td>\n",
              "      <td>99.088342</td>\n",
              "      <td>98.448960</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Random Forest Classifier</td>\n",
              "      <td>97.527129</td>\n",
              "      <td>97.120461</td>\n",
              "      <td>97.958854</td>\n",
              "      <td>97.537856</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                      Model   Accuracy  Precision     Recall   F1 Score\n",
              "0            MLP Classifier  74.573399  70.718270  83.880597  76.739122\n",
              "1           LGBM classifier  98.438824  97.817776  99.088342  98.448960\n",
              "2  Random Forest Classifier  97.527129  97.120461  97.958854  97.537856"
            ]
          },
          "metadata": {},
          "execution_count": 147
        }
      ]
    }
  ]
}