Plagiarism_detection_in_computer_programming_assignments__after_sampling (1).ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "# ***Plagiarism Detection after Sampling process***"
      ],
      "metadata": {
        "id": "MTZL1Ujgl-GY"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "*Initiliasing all the libraries in the beginning*"
      ],
      "metadata": {
        "id": "Fb7pFC6EiaHO"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": true,
        "id": "qgofza2N1n35",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "06a5fda4-b092-4765-822a-cb2cdace8082"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[?25l\r\u001b[K     |██▌                             | 10 kB 29.8 MB/s eta 0:00:01\r\u001b[K     |█████                           | 20 kB 35.7 MB/s eta 0:00:01\r\u001b[K     |███████▍                        | 30 kB 38.8 MB/s eta 0:00:01\r\u001b[K     |█████████▉                      | 40 kB 37.2 MB/s eta 0:00:01\r\u001b[K     |████████████▎                   | 51 kB 26.1 MB/s eta 0:00:01\r\u001b[K     |██████████████▊                 | 61 kB 29.5 MB/s eta 0:00:01\r\u001b[K     |█████████████████▏              | 71 kB 28.8 MB/s eta 0:00:01\r\u001b[K     |███████████████████▋            | 81 kB 30.3 MB/s eta 0:00:01\r\u001b[K     |██████████████████████          | 92 kB 32.1 MB/s eta 0:00:01\r\u001b[K     |████████████████████████▌       | 102 kB 33.3 MB/s eta 0:00:01\r\u001b[K     |███████████████████████████     | 112 kB 33.3 MB/s eta 0:00:01\r\u001b[K     |█████████████████████████████▍  | 122 kB 33.3 MB/s eta 0:00:01\r\u001b[K     |███████████████████████████████▉| 133 kB 33.3 MB/s eta 0:00:01\r\u001b[K     |████████████████████████████████| 133 kB 33.3 MB/s \n",
            "\u001b[?25hLooking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Collecting visualkeras\n",
            "  Downloading visualkeras-0.0.2-py3-none-any.whl (12 kB)\n",
            "Collecting aggdraw>=1.3.11\n",
            "  Downloading aggdraw-1.3.14-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (992 kB)\n",
            "\u001b[K     |████████████████████████████████| 992 kB 24.5 MB/s \n",
            "\u001b[?25hRequirement already satisfied: numpy>=1.18.1 in /usr/local/lib/python3.7/dist-packages (from visualkeras) (1.21.6)\n",
            "Requirement already satisfied: pillow>=6.2.0 in /usr/local/lib/python3.7/dist-packages (from visualkeras) (7.1.2)\n",
            "Installing collected packages: aggdraw, visualkeras\n",
            "Successfully installed aggdraw-1.3.14 visualkeras-0.0.2\n"
          ]
        }
      ],
      "source": [
        "# !pip install fast_ml\n",
        "# from fast_ml.model_development import train_valid_test_split\n",
        "!pip install -q -U keras-tuner\n",
        "!pip install visualkeras"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import sklearn.metrics\n",
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "import keras_tuner as kt\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.utils import shuffle\n",
        "from sklearn.preprocessing import MinMaxScaler\n",
        "import re\n",
        "import imblearn\n",
        "import collections\n",
        "import seaborn as sns"
      ],
      "metadata": {
        "id": "IfbWpsIlHmn2"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "outputs": [],
      "source": [
        "from collections import Counter\n",
        "from imblearn.over_sampling import RandomOverSampler\n",
        "from tensorflow.keras.models import load_model\n",
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Activation, Dense, BatchNormalization,Dropout\n",
        "from tensorflow.keras.optimizers import Adam,SGD\n",
        "from tensorflow.keras import regularizers,layers\n",
        "from tensorflow.keras.callbacks import EarlyStopping,ReduceLROnPlateau,LearningRateScheduler,ModelCheckpoint\n",
        "from tensorflow.keras.losses import BinaryCrossentropy,SparseCategoricalCrossentropy,CategoricalCrossentropy\n",
        "from keras.models import model_from_json\n",
        "from matplotlib import pyplot as plt\n"
      ],
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        },
        "id": "IEAV9wOy1n38"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "***Geting the feature dataset and label using pandas***"
      ],
      "metadata": {
        "id": "pR5FWT3aipMT"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "df=pd.read_csv('/content/sample_data/plagiarism_student_dataset_features.csv',index_col=0)\n",
        "df.index.names=[None]\n",
        "print(df.dtypes)\n",
        "df.describe()\n",
        "sns.set(style = 'whitegrid')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1jYkpMtMMiZg",
        "outputId": "9f29efd4-728a-4ae5-e6d0-c572f00439c3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "total_time               float64\n",
            "compiled                 float64\n",
            "compiled_successfully    float64\n",
            "ran_tests                float64\n",
            "test_results             float64\n",
            "modified                 float64\n",
            "add_lines                float64\n",
            "change                   float64\n",
            "remove_lines             float64\n",
            "cps                      float64\n",
            "paste_lines              float64\n",
            "max_pastes               float64\n",
            "average_test_results     float64\n",
            "avg_paste                float64\n",
            "dtype: object\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "label_A=pd.read_csv('/content/sample_data/label_for_plagiarism_dataset_stat_dynamic.csv',index_col=0)\n",
        "label_A.index.names=[None]\n",
        "label_A.describe()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 300
        },
        "id": "SMY-xjnsM48o",
        "outputId": "3015e927-f4b8-4422-8a78-28aa9d76e494"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "             label\n",
              "count  2339.000000\n",
              "mean      0.284310\n",
              "std       0.451181\n",
              "min       0.000000\n",
              "25%       0.000000\n",
              "50%       0.000000\n",
              "75%       1.000000\n",
              "max       1.000000"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-b3992ac0-592f-43d6-a67b-9a1002585a1f\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>label</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>2339.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>0.284310</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>0.451181</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-b3992ac0-592f-43d6-a67b-9a1002585a1f')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-b3992ac0-592f-43d6-a67b-9a1002585a1f button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-b3992ac0-592f-43d6-a67b-9a1002585a1f');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 270
        },
        "id": "Kelv_VasKkJ8",
        "outputId": "08cbd331-9fc9-4049-e031-80d459e4fbd1"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                   total_time  compiled  compiled_successfully  ran_tests  \\\n",
              "A2016\\student1013   6894752.0    1441.0                 1116.0      689.0   \n",
              "A2016\\student1016     37440.0       NaN                    NaN        NaN   \n",
              "A2016\\student1021   9363152.0    2108.0                 1278.0      787.0   \n",
              "A2016\\student1029   6949656.0    1775.0                 1201.0      852.0   \n",
              "A2016\\student1066   2784368.0     338.0                  223.0      449.0   \n",
              "\n",
              "                   test_results  modified  add_lines   change  remove_lines  \\\n",
              "A2016\\student1013        1349.0   38441.0    12365.0  32597.0        2023.0   \n",
              "A2016\\student1016           NaN       1.0        NaN      NaN           8.0   \n",
              "A2016\\student1021        1605.0   58550.0    19160.0  49003.0        6105.0   \n",
              "A2016\\student1029         754.0   36461.0    15605.0  30332.0        2683.0   \n",
              "A2016\\student1066         810.0   13924.0     6140.0  12169.0         369.0   \n",
              "\n",
              "                        cps  paste_lines  max_pastes  average_test_results  \\\n",
              "A2016\\student1013  0.003436       5202.0       120.0              1.957910   \n",
              "A2016\\student1016  0.000000          NaN         NaN                   NaN   \n",
              "A2016\\student1021  0.003997       7270.0       162.0              2.039390   \n",
              "A2016\\student1029  0.005040       7235.0       165.0              0.884977   \n",
              "A2016\\student1066  0.004970       3988.0       139.0              1.804009   \n",
              "\n",
              "                   avg_paste  \n",
              "A2016\\student1013   4.553441  \n",
              "A2016\\student1016        NaN  \n",
              "A2016\\student1021   5.147455  \n",
              "A2016\\student1029   4.840774  \n",
              "A2016\\student1066   3.470160  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-156e37d5-1ab1-475f-9d57-069a9440a9af\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>total_time</th>\n",
              "      <th>compiled</th>\n",
              "      <th>compiled_successfully</th>\n",
              "      <th>ran_tests</th>\n",
              "      <th>test_results</th>\n",
              "      <th>modified</th>\n",
              "      <th>add_lines</th>\n",
              "      <th>change</th>\n",
              "      <th>remove_lines</th>\n",
              "      <th>cps</th>\n",
              "      <th>paste_lines</th>\n",
              "      <th>max_pastes</th>\n",
              "      <th>average_test_results</th>\n",
              "      <th>avg_paste</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>A2016\\student1013</th>\n",
              "      <td>6894752.0</td>\n",
              "      <td>1441.0</td>\n",
              "      <td>1116.0</td>\n",
              "      <td>689.0</td>\n",
              "      <td>1349.0</td>\n",
              "      <td>38441.0</td>\n",
              "      <td>12365.0</td>\n",
              "      <td>32597.0</td>\n",
              "      <td>2023.0</td>\n",
              "      <td>0.003436</td>\n",
              "      <td>5202.0</td>\n",
              "      <td>120.0</td>\n",
              "      <td>1.957910</td>\n",
              "      <td>4.553441</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>A2016\\student1016</th>\n",
              "      <td>37440.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>1.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>8.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>A2016\\student1021</th>\n",
              "      <td>9363152.0</td>\n",
              "      <td>2108.0</td>\n",
              "      <td>1278.0</td>\n",
              "      <td>787.0</td>\n",
              "      <td>1605.0</td>\n",
              "      <td>58550.0</td>\n",
              "      <td>19160.0</td>\n",
              "      <td>49003.0</td>\n",
              "      <td>6105.0</td>\n",
              "      <td>0.003997</td>\n",
              "      <td>7270.0</td>\n",
              "      <td>162.0</td>\n",
              "      <td>2.039390</td>\n",
              "      <td>5.147455</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>A2016\\student1029</th>\n",
              "      <td>6949656.0</td>\n",
              "      <td>1775.0</td>\n",
              "      <td>1201.0</td>\n",
              "      <td>852.0</td>\n",
              "      <td>754.0</td>\n",
              "      <td>36461.0</td>\n",
              "      <td>15605.0</td>\n",
              "      <td>30332.0</td>\n",
              "      <td>2683.0</td>\n",
              "      <td>0.005040</td>\n",
              "      <td>7235.0</td>\n",
              "      <td>165.0</td>\n",
              "      <td>0.884977</td>\n",
              "      <td>4.840774</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>A2016\\student1066</th>\n",
              "      <td>2784368.0</td>\n",
              "      <td>338.0</td>\n",
              "      <td>223.0</td>\n",
              "      <td>449.0</td>\n",
              "      <td>810.0</td>\n",
              "      <td>13924.0</td>\n",
              "      <td>6140.0</td>\n",
              "      <td>12169.0</td>\n",
              "      <td>369.0</td>\n",
              "      <td>0.004970</td>\n",
              "      <td>3988.0</td>\n",
              "      <td>139.0</td>\n",
              "      <td>1.804009</td>\n",
              "      <td>3.470160</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-156e37d5-1ab1-475f-9d57-069a9440a9af')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-156e37d5-1ab1-475f-9d57-069a9440a9af button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-156e37d5-1ab1-475f-9d57-069a9440a9af');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "***Line Plot of the feature set showing the range of features***"
      ],
      "metadata": {
        "id": "-M4ZZu6_izzO"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "df.plot(figsize=(12,12))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 732
        },
        "id": "kuTLQTFtVH_C",
        "outputId": "2cead543-07da-42c5-8d21-3b6b278d5652"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f748b5f0710>"
            ]
          },
          "metadata": {},
          "execution_count": 7
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 864x864 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "label_A.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "GLVEJSc4KMnG",
        "outputId": "d72c03c6-2bde-4523-8d51-6500d6ad9b37"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                   label\n",
              "A2016\\student1013      0\n",
              "A2016\\student1016      0\n",
              "A2016\\student1021      1\n",
              "A2016\\student1029      1\n",
              "A2016\\student1066      0"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-4688f236-a682-43a0-b5e7-b00917371e6f\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>label</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>A2016\\student1013</th>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>A2016\\student1016</th>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>A2016\\student1021</th>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>A2016\\student1029</th>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>A2016\\student1066</th>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-4688f236-a682-43a0-b5e7-b00917371e6f')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-4688f236-a682-43a0-b5e7-b00917371e6f button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-4688f236-a682-43a0-b5e7-b00917371e6f');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "label_list = label_A['label'].tolist()"
      ],
      "metadata": {
        "id": "yMFaVlkJniHh"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "print(label_list[:10])\n",
        "print(len(label_list))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pbut3_NSoXUc",
        "outputId": "b9de939c-ba60-42a4-cf83-01f8ab31b5a7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[0, 0, 1, 1, 0, 0, 1, 0, 0, 0]\n",
            "2339\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df['label']=label_list\n",
        "df.fillna(0,inplace=True)\n",
        "# df.dropna(inplace = True)\n",
        "trainA=df.iloc[:,:-1]\n",
        "label_A=df.iloc[:,-1]"
      ],
      "metadata": {
        "id": "ZMl6SbE9olOt"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "trainA.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 270
        },
        "id": "_KcFjwlsqf-d",
        "outputId": "8463afb6-60e6-41c7-e069-9c152a8e1ce0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                   total_time  compiled  compiled_successfully  ran_tests  \\\n",
              "A2016\\student1013   6894752.0    1441.0                 1116.0      689.0   \n",
              "A2016\\student1016     37440.0       0.0                    0.0        0.0   \n",
              "A2016\\student1021   9363152.0    2108.0                 1278.0      787.0   \n",
              "A2016\\student1029   6949656.0    1775.0                 1201.0      852.0   \n",
              "A2016\\student1066   2784368.0     338.0                  223.0      449.0   \n",
              "\n",
              "                   test_results  modified  add_lines   change  remove_lines  \\\n",
              "A2016\\student1013        1349.0   38441.0    12365.0  32597.0        2023.0   \n",
              "A2016\\student1016           0.0       1.0        0.0      0.0           8.0   \n",
              "A2016\\student1021        1605.0   58550.0    19160.0  49003.0        6105.0   \n",
              "A2016\\student1029         754.0   36461.0    15605.0  30332.0        2683.0   \n",
              "A2016\\student1066         810.0   13924.0     6140.0  12169.0         369.0   \n",
              "\n",
              "                        cps  paste_lines  max_pastes  average_test_results  \\\n",
              "A2016\\student1013  0.003436       5202.0       120.0              1.957910   \n",
              "A2016\\student1016  0.000000          0.0         0.0              0.000000   \n",
              "A2016\\student1021  0.003997       7270.0       162.0              2.039390   \n",
              "A2016\\student1029  0.005040       7235.0       165.0              0.884977   \n",
              "A2016\\student1066  0.004970       3988.0       139.0              1.804009   \n",
              "\n",
              "                   avg_paste  \n",
              "A2016\\student1013   4.553441  \n",
              "A2016\\student1016   0.000000  \n",
              "A2016\\student1021   5.147455  \n",
              "A2016\\student1029   4.840774  \n",
              "A2016\\student1066   3.470160  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-007ee0b1-9617-49c6-8ac8-366fe36e1805\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>total_time</th>\n",
              "      <th>compiled</th>\n",
              "      <th>compiled_successfully</th>\n",
              "      <th>ran_tests</th>\n",
              "      <th>test_results</th>\n",
              "      <th>modified</th>\n",
              "      <th>add_lines</th>\n",
              "      <th>change</th>\n",
              "      <th>remove_lines</th>\n",
              "      <th>cps</th>\n",
              "      <th>paste_lines</th>\n",
              "      <th>max_pastes</th>\n",
              "      <th>average_test_results</th>\n",
              "      <th>avg_paste</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>A2016\\student1013</th>\n",
              "      <td>6894752.0</td>\n",
              "      <td>1441.0</td>\n",
              "      <td>1116.0</td>\n",
              "      <td>689.0</td>\n",
              "      <td>1349.0</td>\n",
              "      <td>38441.0</td>\n",
              "      <td>12365.0</td>\n",
              "      <td>32597.0</td>\n",
              "      <td>2023.0</td>\n",
              "      <td>0.003436</td>\n",
              "      <td>5202.0</td>\n",
              "      <td>120.0</td>\n",
              "      <td>1.957910</td>\n",
              "      <td>4.553441</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>A2016\\student1016</th>\n",
              "      <td>37440.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>8.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>A2016\\student1021</th>\n",
              "      <td>9363152.0</td>\n",
              "      <td>2108.0</td>\n",
              "      <td>1278.0</td>\n",
              "      <td>787.0</td>\n",
              "      <td>1605.0</td>\n",
              "      <td>58550.0</td>\n",
              "      <td>19160.0</td>\n",
              "      <td>49003.0</td>\n",
              "      <td>6105.0</td>\n",
              "      <td>0.003997</td>\n",
              "      <td>7270.0</td>\n",
              "      <td>162.0</td>\n",
              "      <td>2.039390</td>\n",
              "      <td>5.147455</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>A2016\\student1029</th>\n",
              "      <td>6949656.0</td>\n",
              "      <td>1775.0</td>\n",
              "      <td>1201.0</td>\n",
              "      <td>852.0</td>\n",
              "      <td>754.0</td>\n",
              "      <td>36461.0</td>\n",
              "      <td>15605.0</td>\n",
              "      <td>30332.0</td>\n",
              "      <td>2683.0</td>\n",
              "      <td>0.005040</td>\n",
              "      <td>7235.0</td>\n",
              "      <td>165.0</td>\n",
              "      <td>0.884977</td>\n",
              "      <td>4.840774</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>A2016\\student1066</th>\n",
              "      <td>2784368.0</td>\n",
              "      <td>338.0</td>\n",
              "      <td>223.0</td>\n",
              "      <td>449.0</td>\n",
              "      <td>810.0</td>\n",
              "      <td>13924.0</td>\n",
              "      <td>6140.0</td>\n",
              "      <td>12169.0</td>\n",
              "      <td>369.0</td>\n",
              "      <td>0.004970</td>\n",
              "      <td>3988.0</td>\n",
              "      <td>139.0</td>\n",
              "      <td>1.804009</td>\n",
              "      <td>3.470160</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-007ee0b1-9617-49c6-8ac8-366fe36e1805')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-007ee0b1-9617-49c6-8ac8-366fe36e1805 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-007ee0b1-9617-49c6-8ac8-366fe36e1805');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "label_A.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "f0Z_i-rTTSsz",
        "outputId": "fcea100e-5495-4151-c32d-9fde86ce994c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "A2016\\student1013    0\n",
              "A2016\\student1016    0\n",
              "A2016\\student1021    1\n",
              "A2016\\student1029    1\n",
              "A2016\\student1066    0\n",
              "Name: label, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "***The count of plagiarised and not plagiarised dataset available to us which is skewed, as the count of not plagiarised is greater than the count of plagiarised***"
      ],
      "metadata": {
        "id": "VSU4uoVSjn0Q"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "sns.countplot(label_A)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 359
        },
        "id": "RVOE9wMEUc1G",
        "outputId": "6244f2d4-4c27-44fe-b563-a409a38cd637"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/seaborn/_decorators.py:43: FutureWarning: Pass the following variable as a keyword arg: x. From version 0.12, the only valid positional argument will be `data`, and passing other arguments without an explicit keyword will result in an error or misinterpretation.\n",
            "  FutureWarning\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f748b3a6fd0>"
            ]
          },
          "metadata": {},
          "execution_count": 14
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "***Scaling the values of features between 0 to 1 for better results and classification***"
      ],
      "metadata": {
        "id": "ccY9UZCDj_4c"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "scaler=MinMaxScaler(feature_range=(0,1))\n",
        "train_A=scaler.fit_transform(trainA)"
      ],
      "metadata": {
        "id": "uPkJ8hmRIfxb"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(train_A.dtype)\n",
        "print(train_A.shape)\n",
        "print(train_A)"
      ],
      "metadata": {
        "id": "DXtI8bfLKoxn",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "2b329a21-30c4-40f7-b6c3-1fd3140a678f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "float64\n",
            "(2339, 14)\n",
            "[[3.00070017e-01 1.34384034e-01 1.52417372e-01 ... 6.00000000e-03\n",
            "  1.95791002e-01 1.05267808e-02]\n",
            " [1.62944533e-03 0.00000000e+00 0.00000000e+00 ... 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00]\n",
            " [4.07498512e-01 1.96586776e-01 1.74542475e-01 ... 8.10000000e-03\n",
            "  2.03939009e-01 1.19000408e-02]\n",
            " ...\n",
            " [5.37605544e-01 5.28676676e-01 4.87298552e-01 ... 1.16000000e-02\n",
            "  6.28982528e-02 1.11948382e-02]\n",
            " [3.69758748e-04 0.00000000e+00 0.00000000e+00 ... 1.00000000e-04\n",
            "  0.00000000e+00 3.39974979e-03]\n",
            " [2.87237439e-01 1.54714166e-01 1.35618683e-01 ... 1.62500000e-02\n",
            "  1.02476780e-01 1.58131852e-02]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(train_A)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 518
        },
        "id": "_p4WkAOte_x7",
        "outputId": "abe823c5-6b02-4a6a-c8a2-20238d9cc308"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[<matplotlib.lines.Line2D at 0x7f7489547190>,\n",
              " <matplotlib.lines.Line2D at 0x7f74895473d0>,\n",
              " <matplotlib.lines.Line2D at 0x7f7489547590>,\n",
              " <matplotlib.lines.Line2D at 0x7f7489547750>,\n",
              " <matplotlib.lines.Line2D at 0x7f7489547910>,\n",
              " <matplotlib.lines.Line2D at 0x7f7489547ad0>,\n",
              " <matplotlib.lines.Line2D at 0x7f7489547d10>,\n",
              " <matplotlib.lines.Line2D at 0x7f7489547ed0>,\n",
              " <matplotlib.lines.Line2D at 0x7f74895478d0>,\n",
              " <matplotlib.lines.Line2D at 0x7f7489547b10>,\n",
              " <matplotlib.lines.Line2D at 0x7f74895471d0>,\n",
              " <matplotlib.lines.Line2D at 0x7f748954d550>,\n",
              " <matplotlib.lines.Line2D at 0x7f748954d710>,\n",
              " <matplotlib.lines.Line2D at 0x7f748954d8d0>]"
            ]
          },
          "metadata": {},
          "execution_count": 17
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "label_A=label_A.to_numpy()\n",
        "print(label_A.dtype)\n",
        "# label_A=label_A.astype(float)\n",
        "print(label_A.shape)\n",
        "print(label_A)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "sfF641gqM4JQ",
        "outputId": "c0c8db28-6764-4be1-d604-22dde5260fdc"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "int64\n",
            "(2339,)\n",
            "[0 0 1 ... 0 0 0]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "***Splitting the dataset into training and testing with a train_size split of 0.8***"
      ],
      "metadata": {
        "id": "tiOqjf4VkMfo"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "x_train,x_test,y_train,y_test= train_test_split(train_A,label_A,train_size=0.8,shuffle=True)"
      ],
      "metadata": {
        "id": "i7l9ytP3TdYY"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(x_train.shape)\n",
        "print(y_train.shape)\n",
        "print(x_test.shape)\n",
        "print(y_test.shape)\n",
        "print(sorted(Counter(y_train).items()))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZCm49h-JajoB",
        "outputId": "404277e1-13f9-418a-912c-4a10dc9576ec"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(1871, 14)\n",
            "(1871,)\n",
            "(468, 14)\n",
            "(468,)\n",
            "[(0, 1342), (1, 529)]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "smote=imblearn.over_sampling.SMOTE(sampling_strategy='minority')\n",
        "x_train_resampled, y_train_resampled = smote.fit_resample(x_train, y_train)\n",
        "print(sorted(Counter(y_train_resampled).items()))"
      ],
      "metadata": {
        "id": "N6p6nDYiPsMc",
        "outputId": "6be97886-eacc-4ac7-934f-c8a68d7863ac",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[(0, 1341), (1, 1341)]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "***As the dataset is skewed we sample it using the methods of both under sampling and over sampling which is achieved through smote enn and the sampling strategy we chose is minority***"
      ],
      "metadata": {
        "id": "yuJuDwVMkW2e"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "smote_enn = imblearn.combine.SMOTEENN(random_state=0,sampling_strategy='minority')\n",
        "x_train_resampled, y_train_resampled = smote_enn.fit_resample(x_train, y_train)\n",
        "print(sorted(Counter(y_train_resampled).items()))"
      ],
      "metadata": {
        "id": "0JmL9bX6HOYw",
        "outputId": "aceb848a-4b6f-4218-ee1a-25b5bbc48ce4",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[(0, 703), (1, 956)]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "***This is the countplot after the sampling is performed***"
      ],
      "metadata": {
        "id": "vmAiOnLQlY_K"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "sns.countplot(y_train_resampled)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 345
        },
        "id": "i18qLXL-01_f",
        "outputId": "abc235e6-6191-4b3a-dcaf-d64967ad57fc"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/seaborn/_decorators.py:43: FutureWarning: Pass the following variable as a keyword arg: x. From version 0.12, the only valid positional argument will be `data`, and passing other arguments without an explicit keyword will result in an error or misinterpretation.\n",
            "  FutureWarning\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f74894bf350>"
            ]
          },
          "metadata": {},
          "execution_count": 22
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "***The callbacks are defined here to stop training the model when there is no change in the parameter we wish to monitor***"
      ],
      "metadata": {
        "id": "alZxG6d-ledD"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# callbacks\n",
        "# early stopping\n",
        "custom_early_stopping = EarlyStopping(\n",
        "    monitor='val_accuracy', \n",
        "    patience=5, \n",
        "    min_delta=0, \n",
        "    mode='max',\n",
        "    restore_best_weights=True\n",
        "\n",
        ")\n",
        "custom_early_stopping_loss = EarlyStopping(\n",
        "    monitor='val_loss', \n",
        "    patience=5, \n",
        "    min_delta=0, \n",
        "    mode='min',\n",
        "    restore_best_weights=True\n",
        ")\n",
        "custom_early_stopping_recall=EarlyStopping(\n",
        "    monitor='val_recall', \n",
        "    patience=5, \n",
        "    min_delta=0, \n",
        "    mode='max',\n",
        "    restore_best_weights=True\n",
        ")\n",
        "\n",
        "# reduce_lr callback\n",
        "reduce_lr = ReduceLROnPlateau(\n",
        "    monitor='val_accuracy', \n",
        "    factor=0.2,   \n",
        "    patience=0, \n",
        "    mode=\"max\",\n",
        "    verbose=2\n",
        ")\n",
        "# lr_decay callback\n",
        "def lr_decay(epoch, lr):\n",
        "    if epoch != 0 and epoch % 5 == 0:\n",
        "        return lr * 0.2\n",
        "    return lr\n",
        "\n",
        "# modelcheckpoint callback\n",
        "filepath = 'my_best_model_acc{val_accuracy:.2f}.hdf5'\n",
        "checkpoint = ModelCheckpoint(filepath=filepath, \n",
        "                             monitor='val_accuracy',\n",
        "                             verbose=1, \n",
        "                             save_best_only=True,\n",
        "                             mode='max')\n",
        "# callbacks = [checkpoint]"
      ],
      "metadata": {
        "id": "a7j4MOOzm8d2"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "***A sequential model is built in the function using keras tuner which selects the best model for the inputs based on the range of layers given(2-40) and number of neurons between 16-512 with a step size of 32 and also the learning rate. It selects the best hyperparameters based on the input dataset provided***"
      ],
      "metadata": {
        "id": "ToBLESILlrx8"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "outputs": [],
      "source": [
        "def build_model(hp):\n",
        "  model=keras.Sequential()\n",
        "  \n",
        "  for layer in range(hp.Int('num_layers',2,40)):\n",
        "    model.add(layers.Dense(units=hp.Int('units' + str(layer),\n",
        "                                            min_value=16,\n",
        "                                            max_value=512,\n",
        "                                            step=32),\n",
        "                               activation='relu')),\n",
        "  model.add(layers.Dense(units=1,activation='sigmoid'))\n",
        "  model.compile(optimizer=SGD(hp.Choice('learning_rate',[1e-2, 1e-3, 1e-4,1e-5,1e-6,1e-7,1e-8])),\n",
        "                              loss=keras.losses.BinaryCrossentropy(),metrics=['accuracy',tf.keras.metrics.Precision(),tf.keras.metrics.Recall()])\n",
        "  return model "
      ],
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        },
        "id": "CJ_0REee1n39"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "tuner = kt.RandomSearch(\n",
        "    build_model,\n",
        "    objective=kt.Objective(\"val_accuracy\",direction=\"max\"),\n",
        "    max_trials=4,\n",
        "    executions_per_trial=5,\n",
        "    directory='project2',\n",
        "    overwrite=True,\n",
        "    project_name='Plagiarism Detection')"
      ],
      "metadata": {
        "id": "VkgKIzCT2d6j"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "tuner.search_space_summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1QCIvQRHbvyy",
        "outputId": "49c2a81c-17a0-4246-f4eb-082c0df2ad32"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Search space summary\n",
            "Default search space size: 4\n",
            "num_layers (Int)\n",
            "{'default': None, 'conditions': [], 'min_value': 2, 'max_value': 40, 'step': 1, 'sampling': None}\n",
            "units0 (Int)\n",
            "{'default': None, 'conditions': [], 'min_value': 16, 'max_value': 512, 'step': 32, 'sampling': None}\n",
            "units1 (Int)\n",
            "{'default': None, 'conditions': [], 'min_value': 16, 'max_value': 512, 'step': 32, 'sampling': None}\n",
            "learning_rate (Choice)\n",
            "{'default': 0.01, 'conditions': [], 'values': [0.01, 0.001, 0.0001, 1e-05, 1e-06, 1e-07, 1e-08], 'ordered': True}\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "tuner.search(x_train_resampled,\n",
        "          y_train_resampled,\n",
        "          epochs=100,\n",
        "          validation_split=0.25,\n",
        "          callbacks=[\n",
        "                     custom_early_stopping,\n",
        "                    reduce_lr]\n",
        "                )"
      ],
      "metadata": {
        "id": "MncBvyyUAjya",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "15810bba-ad66-4530-e3c5-86afeee8ebf8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Trial 4 Complete [00h 01m 02s]\n",
            "val_accuracy: 0.0\n",
            "\n",
            "Best val_accuracy So Far: 0.46066349744796753\n",
            "Total elapsed time: 00h 02m 41s\n",
            "INFO:tensorflow:Oracle triggered exit\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "tuner.results_summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "U9cjk4cbeuk7",
        "outputId": "c506ac16-643f-4566-ba66-c1e1aeaa2369"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Results summary\n",
            "Results in project2/Plagiarism Detection\n",
            "Showing 10 best trials\n",
            "<keras_tuner.engine.objective.Objective object at 0x7f16bb9af510>\n",
            "Trial summary\n",
            "Hyperparameters:\n",
            "num_layers: 3\n",
            "units0: 112\n",
            "units1: 112\n",
            "learning_rate: 0.001\n",
            "units2: 240\n",
            "units3: 176\n",
            "units4: 304\n",
            "units5: 208\n",
            "units6: 464\n",
            "units7: 16\n",
            "units8: 368\n",
            "units9: 144\n",
            "units10: 336\n",
            "units11: 496\n",
            "units12: 464\n",
            "units13: 432\n",
            "units14: 272\n",
            "units15: 272\n",
            "units16: 176\n",
            "units17: 272\n",
            "units18: 144\n",
            "units19: 368\n",
            "units20: 368\n",
            "units21: 48\n",
            "units22: 368\n",
            "units23: 368\n",
            "units24: 208\n",
            "units25: 400\n",
            "units26: 496\n",
            "units27: 144\n",
            "units28: 368\n",
            "units29: 112\n",
            "units30: 208\n",
            "units31: 112\n",
            "units32: 112\n",
            "units33: 432\n",
            "units34: 368\n",
            "units35: 304\n",
            "Score: 0.46066349744796753\n",
            "Trial summary\n",
            "Hyperparameters:\n",
            "num_layers: 36\n",
            "units0: 240\n",
            "units1: 272\n",
            "learning_rate: 1e-08\n",
            "units2: 16\n",
            "units3: 16\n",
            "units4: 16\n",
            "units5: 16\n",
            "units6: 16\n",
            "units7: 16\n",
            "units8: 16\n",
            "units9: 16\n",
            "units10: 16\n",
            "units11: 16\n",
            "units12: 16\n",
            "units13: 16\n",
            "units14: 16\n",
            "units15: 16\n",
            "units16: 16\n",
            "units17: 16\n",
            "units18: 16\n",
            "units19: 16\n",
            "units20: 16\n",
            "units21: 16\n",
            "units22: 16\n",
            "units23: 16\n",
            "units24: 16\n",
            "units25: 16\n",
            "units26: 16\n",
            "units27: 16\n",
            "units28: 16\n",
            "units29: 16\n",
            "units30: 16\n",
            "units31: 16\n",
            "units32: 16\n",
            "units33: 16\n",
            "units34: 16\n",
            "units35: 16\n",
            "Score: 0.0\n",
            "Trial summary\n",
            "Hyperparameters:\n",
            "num_layers: 30\n",
            "units0: 208\n",
            "units1: 208\n",
            "learning_rate: 0.001\n",
            "units2: 272\n",
            "units3: 400\n",
            "units4: 80\n",
            "units5: 496\n",
            "units6: 336\n",
            "units7: 208\n",
            "units8: 240\n",
            "units9: 80\n",
            "units10: 176\n",
            "units11: 432\n",
            "units12: 176\n",
            "units13: 208\n",
            "units14: 16\n",
            "units15: 368\n",
            "units16: 112\n",
            "units17: 80\n",
            "units18: 208\n",
            "units19: 80\n",
            "units20: 464\n",
            "units21: 464\n",
            "units22: 464\n",
            "units23: 112\n",
            "units24: 80\n",
            "units25: 144\n",
            "units26: 432\n",
            "units27: 208\n",
            "units28: 208\n",
            "units29: 112\n",
            "units30: 464\n",
            "units31: 112\n",
            "units32: 464\n",
            "units33: 144\n",
            "units34: 48\n",
            "units35: 240\n",
            "Score: 0.0\n",
            "Trial summary\n",
            "Hyperparameters:\n",
            "num_layers: 28\n",
            "units0: 496\n",
            "units1: 240\n",
            "learning_rate: 1e-05\n",
            "units2: 208\n",
            "units3: 464\n",
            "units4: 496\n",
            "units5: 240\n",
            "units6: 464\n",
            "units7: 272\n",
            "units8: 112\n",
            "units9: 272\n",
            "units10: 240\n",
            "units11: 368\n",
            "units12: 48\n",
            "units13: 400\n",
            "units14: 400\n",
            "units15: 432\n",
            "units16: 304\n",
            "units17: 112\n",
            "units18: 432\n",
            "units19: 208\n",
            "units20: 144\n",
            "units21: 112\n",
            "units22: 496\n",
            "units23: 208\n",
            "units24: 240\n",
            "units25: 176\n",
            "units26: 80\n",
            "units27: 496\n",
            "units28: 16\n",
            "units29: 368\n",
            "units30: 432\n",
            "units31: 464\n",
            "units32: 464\n",
            "units33: 496\n",
            "units34: 336\n",
            "units35: 432\n",
            "Score: 0.0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "***The model with best validation accuracy is chosen after searching it using RandomSearch***"
      ],
      "metadata": {
        "id": "ZShyMUFwnaaY"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "best_hps=tuner.get_best_hyperparameters(num_trials=1)[0]\n",
        "print(best_hps.get('num_layers'))\n",
        "print(best_hps.get('learning_rate'))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NDfe9gexgAzL",
        "outputId": "8d8d43ae-6d70-40e1-e44f-907600d6e763"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "3\n",
            "0.001\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "***The best model is built and fit on the dataset. The model is also retrained to get better accuracy and recall values***"
      ],
      "metadata": {
        "id": "pWG9Kkz9nzRo"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "model = tuner.hypermodel.build(best_hps)\n",
        "history = model.fit(x_train_resampled, y_train_resampled, epochs=100, validation_split=0.25,callbacks=[custom_early_stopping,reduce_lr])\n",
        "# val_acc_per_epoch=history.history['val_loss']\n",
        "val_acc_per_epoch = history.history['val_accuracy']\n",
        "best_epoch = val_acc_per_epoch.index(max(val_acc_per_epoch)) + 1\n",
        "print('Best epoch: %d' % (best_epoch,))"
      ],
      "metadata": {
        "id": "TOGna7WQPHYR",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "1e5b5f48-9e33-40f4-884a-549fc7846b26"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "40/40 [==============================] - 1s 13ms/step - loss: 0.6975 - accuracy: 0.5685 - precision_1: 0.0000e+00 - recall_1: 0.0000e+00 - val_loss: 0.7112 - val_accuracy: 0.0000e+00 - val_precision_1: 0.0000e+00 - val_recall_1: 0.0000e+00 - lr: 0.0010\n",
            "Epoch 2/100\n",
            "40/40 [==============================] - ETA: 0s - loss: 0.6959 - accuracy: 0.5685 - precision_1: 0.0000e+00 - recall_1: 0.0000e+00\n",
            "Epoch 2: ReduceLROnPlateau reducing learning rate to 0.00020000000949949026.\n",
            "40/40 [==============================] - 0s 5ms/step - loss: 0.6959 - accuracy: 0.5685 - precision_1: 0.0000e+00 - recall_1: 0.0000e+00 - val_loss: 0.7117 - val_accuracy: 0.0000e+00 - val_precision_1: 0.0000e+00 - val_recall_1: 0.0000e+00 - lr: 0.0010\n",
            "Epoch 3/100\n",
            "22/40 [===============>..............] - ETA: 0s - loss: 0.6959 - accuracy: 0.5384 - precision_1: 0.0000e+00 - recall_1: 0.0000e+00\n",
            "Epoch 3: ReduceLROnPlateau reducing learning rate to 4.0000001899898055e-05.\n",
            "40/40 [==============================] - 0s 5ms/step - loss: 0.6948 - accuracy: 0.5685 - precision_1: 0.0000e+00 - recall_1: 0.0000e+00 - val_loss: 0.7117 - val_accuracy: 0.0000e+00 - val_precision_1: 0.0000e+00 - val_recall_1: 0.0000e+00 - lr: 2.0000e-04\n",
            "Epoch 4/100\n",
            "39/40 [============================>.] - ETA: 0s - loss: 0.6946 - accuracy: 0.5681 - precision_1: 0.0000e+00 - recall_1: 0.0000e+00\n",
            "Epoch 4: ReduceLROnPlateau reducing learning rate to 8.000000525498762e-06.\n",
            "40/40 [==============================] - 0s 4ms/step - loss: 0.6946 - accuracy: 0.5685 - precision_1: 0.0000e+00 - recall_1: 0.0000e+00 - val_loss: 0.7117 - val_accuracy: 0.0000e+00 - val_precision_1: 0.0000e+00 - val_recall_1: 0.0000e+00 - lr: 4.0000e-05\n",
            "Epoch 5/100\n",
            "33/40 [=======================>......] - ETA: 0s - loss: 0.6947 - accuracy: 0.5634 - precision_1: 0.0000e+00 - recall_1: 0.0000e+00\n",
            "Epoch 5: ReduceLROnPlateau reducing learning rate to 1.6000001778593287e-06.\n",
            "40/40 [==============================] - 0s 5ms/step - loss: 0.6946 - accuracy: 0.5685 - precision_1: 0.0000e+00 - recall_1: 0.0000e+00 - val_loss: 0.7117 - val_accuracy: 0.0000e+00 - val_precision_1: 0.0000e+00 - val_recall_1: 0.0000e+00 - lr: 8.0000e-06\n",
            "Epoch 6/100\n",
            "37/40 [==========================>...] - ETA: 0s - loss: 0.6945 - accuracy: 0.5701 - precision_1: 0.0000e+00 - recall_1: 0.0000e+00\n",
            "Epoch 6: ReduceLROnPlateau reducing learning rate to 3.200000264769187e-07.\n",
            "40/40 [==============================] - 0s 4ms/step - loss: 0.6945 - accuracy: 0.5685 - precision_1: 0.0000e+00 - recall_1: 0.0000e+00 - val_loss: 0.7117 - val_accuracy: 0.0000e+00 - val_precision_1: 0.0000e+00 - val_recall_1: 0.0000e+00 - lr: 1.6000e-06\n",
            "Best epoch: 1\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "hypermodel = tuner.hypermodel.build(best_hps)\n",
        "\n",
        "# Retrain the model\n",
        "hypermodel.fit(x_train_resampled, y_train_resampled, epochs=best_epoch, validation_split=0.2,callbacks=[reduce_lr,custom_early_stopping])"
      ],
      "metadata": {
        "id": "dx2Pfc8u4mFG",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "e656100b-13b6-4b4a-dcec-531769d41fa5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "43/43 [==============================] - 2s 14ms/step - loss: 0.6869 - accuracy: 0.7470 - precision_2: 0.6640 - recall_2: 0.9286 - val_loss: 0.6766 - val_accuracy: 0.9318 - val_precision_2: 1.0000 - val_recall_2: 0.9318 - lr: 0.0010\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7f16b7b00dd0>"
            ]
          },
          "metadata": {},
          "execution_count": 222
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from statistics import median, mode\n",
        "eval_result = hypermodel.evaluate(x_test, y_test)\n",
        "eval_result_train=hypermodel.evaluate(x_train_resampled,y_train_resampled)\n",
        "eval_result\n",
        "print(\"[test loss, test accuracy,test precision,test recall]:\", eval_result)# print(predictions.dtype)\n",
        "predictions=hypermodel.predict(x_test)\n",
        "\n",
        "print(sum(predictions)/len(predictions))\n",
        "print(median(predictions))\n",
        "val=min(sum(predictions)/len(predictions),median(predictions))\n",
        "predictions"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hUO07Hryv2_1",
        "outputId": "6a463fff-5bb4-41a8-d77d-e77aa647a6ec"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "15/15 [==============================] - 0s 2ms/step - loss: 0.6954 - accuracy: 0.5406 - precision_2: 0.3788 - recall_2: 0.9259\n",
            "53/53 [==============================] - 0s 2ms/step - loss: 0.6844 - accuracy: 0.8273 - precision_2: 0.8012 - recall_2: 0.9297\n",
            "[test loss, test accuracy,test precision,test recall]: [0.6954474449157715, 0.5405982732772827, 0.3787878751754761, 0.9259259104728699]\n",
            "[0.5059709]\n",
            "[0.50354123]\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0.51376194],\n",
              "       [0.50781643],\n",
              "       [0.5027408 ],\n",
              "       [0.4990764 ],\n",
              "       [0.5053578 ],\n",
              "       [0.5028323 ],\n",
              "       [0.4991333 ],\n",
              "       [0.5327133 ],\n",
              "       [0.50258887],\n",
              "       [0.5285369 ],\n",
              "       [0.5080836 ],\n",
              "       [0.5044049 ],\n",
              "       [0.49912548],\n",
              "       [0.5117953 ],\n",
              "       [0.501682  ],\n",
              "       [0.4990761 ],\n",
              "       [0.50270665],\n",
              "       [0.49936494],\n",
              "       [0.5092619 ],\n",
              "       [0.49907503],\n",
              "       [0.5114746 ],\n",
              "       [0.50105184],\n",
              "       [0.5078654 ],\n",
              "       [0.49919617],\n",
              "       [0.4990785 ],\n",
              "       [0.49911734],\n",
              "       [0.5011914 ],\n",
              "       [0.5137399 ],\n",
              "       [0.51057553],\n",
              "       [0.5119382 ],\n",
              "       [0.5009475 ],\n",
              "       [0.50524575],\n",
              "       [0.50150293],\n",
              "       [0.49907407],\n",
              "       [0.5207939 ],\n",
              "       [0.5091077 ],\n",
              "       [0.5024663 ],\n",
              "       [0.50918067],\n",
              "       [0.4991221 ],\n",
              "       [0.4990741 ],\n",
              "       [0.49912933],\n",
              "       [0.5132318 ],\n",
              "       [0.4992469 ],\n",
              "       [0.5015606 ],\n",
              "       [0.5010841 ],\n",
              "       [0.49907595],\n",
              "       [0.49907467],\n",
              "       [0.5118811 ],\n",
              "       [0.51330143],\n",
              "       [0.50914544],\n",
              "       [0.5058203 ],\n",
              "       [0.49965233],\n",
              "       [0.5013863 ],\n",
              "       [0.5048499 ],\n",
              "       [0.49907663],\n",
              "       [0.5078607 ],\n",
              "       [0.4990743 ],\n",
              "       [0.49992654],\n",
              "       [0.50840306],\n",
              "       [0.50606483],\n",
              "       [0.49986053],\n",
              "       [0.49907452],\n",
              "       [0.5155242 ],\n",
              "       [0.5125266 ],\n",
              "       [0.49913654],\n",
              "       [0.4995216 ],\n",
              "       [0.5098537 ],\n",
              "       [0.5003489 ],\n",
              "       [0.50157505],\n",
              "       [0.50716794],\n",
              "       [0.5028097 ],\n",
              "       [0.49917018],\n",
              "       [0.5097839 ],\n",
              "       [0.5150134 ],\n",
              "       [0.5135348 ],\n",
              "       [0.49907735],\n",
              "       [0.50383705],\n",
              "       [0.5054582 ],\n",
              "       [0.5107724 ],\n",
              "       [0.5003936 ],\n",
              "       [0.5006452 ],\n",
              "       [0.5072338 ],\n",
              "       [0.4999558 ],\n",
              "       [0.5116803 ],\n",
              "       [0.5270769 ],\n",
              "       [0.51913124],\n",
              "       [0.5024555 ],\n",
              "       [0.49905467],\n",
              "       [0.50976646],\n",
              "       [0.49907452],\n",
              "       [0.52243567],\n",
              "       [0.5169157 ],\n",
              "       [0.5091919 ],\n",
              "       [0.49812463],\n",
              "       [0.50290793],\n",
              "       [0.50303584],\n",
              "       [0.5070553 ],\n",
              "       [0.49903727],\n",
              "       [0.49907434],\n",
              "       [0.50260264],\n",
              "       [0.50747997],\n",
              "       [0.5151403 ],\n",
              "       [0.501888  ],\n",
              "       [0.50012296],\n",
              "       [0.507975  ],\n",
              "       [0.5059931 ],\n",
              "       [0.5057707 ],\n",
              "       [0.49907446],\n",
              "       [0.5198322 ],\n",
              "       [0.4990815 ],\n",
              "       [0.51884496],\n",
              "       [0.51697564],\n",
              "       [0.52086204],\n",
              "       [0.51092327],\n",
              "       [0.49907464],\n",
              "       [0.5032434 ],\n",
              "       [0.5078906 ],\n",
              "       [0.51487726],\n",
              "       [0.49948722],\n",
              "       [0.50135803],\n",
              "       [0.49889794],\n",
              "       [0.51679015],\n",
              "       [0.5269275 ],\n",
              "       [0.50585586],\n",
              "       [0.4991291 ],\n",
              "       [0.5036365 ],\n",
              "       [0.5133368 ],\n",
              "       [0.5042591 ],\n",
              "       [0.5102703 ],\n",
              "       [0.5011416 ],\n",
              "       [0.510916  ],\n",
              "       [0.5145301 ],\n",
              "       [0.5050794 ],\n",
              "       [0.4990844 ],\n",
              "       [0.49907452],\n",
              "       [0.5115307 ],\n",
              "       [0.5156166 ],\n",
              "       [0.510751  ],\n",
              "       [0.507657  ],\n",
              "       [0.5043342 ],\n",
              "       [0.5141082 ],\n",
              "       [0.50357705],\n",
              "       [0.49907818],\n",
              "       [0.49968612],\n",
              "       [0.49963826],\n",
              "       [0.5064788 ],\n",
              "       [0.50566196],\n",
              "       [0.50781393],\n",
              "       [0.5005422 ],\n",
              "       [0.49990532],\n",
              "       [0.4998944 ],\n",
              "       [0.51154464],\n",
              "       [0.49907452],\n",
              "       [0.4997467 ],\n",
              "       [0.5062709 ],\n",
              "       [0.50949734],\n",
              "       [0.50538725],\n",
              "       [0.51641583],\n",
              "       [0.5004813 ],\n",
              "       [0.5083211 ],\n",
              "       [0.49887803],\n",
              "       [0.5125399 ],\n",
              "       [0.51002336],\n",
              "       [0.51188815],\n",
              "       [0.5072596 ],\n",
              "       [0.49907663],\n",
              "       [0.50874716],\n",
              "       [0.5000236 ],\n",
              "       [0.5068715 ],\n",
              "       [0.5002735 ],\n",
              "       [0.50083524],\n",
              "       [0.50354296],\n",
              "       [0.51637113],\n",
              "       [0.5047509 ],\n",
              "       [0.50822806],\n",
              "       [0.49909884],\n",
              "       [0.50925356],\n",
              "       [0.50239027],\n",
              "       [0.5001804 ],\n",
              "       [0.5074072 ],\n",
              "       [0.5142662 ],\n",
              "       [0.5076378 ],\n",
              "       [0.50792295],\n",
              "       [0.49911353],\n",
              "       [0.5231765 ],\n",
              "       [0.5044648 ],\n",
              "       [0.49923345],\n",
              "       [0.5000966 ],\n",
              "       [0.5246545 ],\n",
              "       [0.4991906 ],\n",
              "       [0.5073056 ],\n",
              "       [0.5160626 ],\n",
              "       [0.499298  ],\n",
              "       [0.50981814],\n",
              "       [0.5120668 ],\n",
              "       [0.49977034],\n",
              "       [0.5090322 ],\n",
              "       [0.5174388 ],\n",
              "       [0.5003018 ],\n",
              "       [0.5020799 ],\n",
              "       [0.51450247],\n",
              "       [0.5190267 ],\n",
              "       [0.53128326],\n",
              "       [0.5027911 ],\n",
              "       [0.49907452],\n",
              "       [0.49931586],\n",
              "       [0.5239549 ],\n",
              "       [0.5060761 ],\n",
              "       [0.5046585 ],\n",
              "       [0.49931693],\n",
              "       [0.49969575],\n",
              "       [0.4991131 ],\n",
              "       [0.49923548],\n",
              "       [0.507625  ],\n",
              "       [0.4990743 ],\n",
              "       [0.49912265],\n",
              "       [0.5168246 ],\n",
              "       [0.5159448 ],\n",
              "       [0.5134185 ],\n",
              "       [0.51394945],\n",
              "       [0.5029661 ],\n",
              "       [0.5061811 ],\n",
              "       [0.51083195],\n",
              "       [0.4998783 ],\n",
              "       [0.50736886],\n",
              "       [0.49907622],\n",
              "       [0.506901  ],\n",
              "       [0.49913162],\n",
              "       [0.5132446 ],\n",
              "       [0.499074  ],\n",
              "       [0.499408  ],\n",
              "       [0.50212795],\n",
              "       [0.4990743 ],\n",
              "       [0.49914777],\n",
              "       [0.49932218],\n",
              "       [0.50164205],\n",
              "       [0.50614107],\n",
              "       [0.51172745],\n",
              "       [0.5134094 ],\n",
              "       [0.5125656 ],\n",
              "       [0.5134886 ],\n",
              "       [0.5113642 ],\n",
              "       [0.50284785],\n",
              "       [0.50255513],\n",
              "       [0.49912202],\n",
              "       [0.49907583],\n",
              "       [0.5110425 ],\n",
              "       [0.50117725],\n",
              "       [0.50894886],\n",
              "       [0.5041742 ],\n",
              "       [0.51609236],\n",
              "       [0.50887865],\n",
              "       [0.5051174 ],\n",
              "       [0.4990779 ],\n",
              "       [0.5029797 ],\n",
              "       [0.5002609 ],\n",
              "       [0.5050168 ],\n",
              "       [0.4993945 ],\n",
              "       [0.5132862 ],\n",
              "       [0.5269445 ],\n",
              "       [0.5114525 ],\n",
              "       [0.5079261 ],\n",
              "       [0.5138318 ],\n",
              "       [0.49925375],\n",
              "       [0.4991814 ],\n",
              "       [0.4999311 ],\n",
              "       [0.5198304 ],\n",
              "       [0.49907452],\n",
              "       [0.49907425],\n",
              "       [0.5158931 ],\n",
              "       [0.50645304],\n",
              "       [0.499874  ],\n",
              "       [0.5150359 ],\n",
              "       [0.49915063],\n",
              "       [0.5073296 ],\n",
              "       [0.5086526 ],\n",
              "       [0.51114565],\n",
              "       [0.5006922 ],\n",
              "       [0.5154217 ],\n",
              "       [0.50328577],\n",
              "       [0.4994871 ],\n",
              "       [0.51495147],\n",
              "       [0.50646937],\n",
              "       [0.4998505 ],\n",
              "       [0.5004957 ],\n",
              "       [0.51854277],\n",
              "       [0.49997157],\n",
              "       [0.49912035],\n",
              "       [0.49909285],\n",
              "       [0.49907404],\n",
              "       [0.49929687],\n",
              "       [0.49909082],\n",
              "       [0.5049586 ],\n",
              "       [0.51062745],\n",
              "       [0.50682664],\n",
              "       [0.5163666 ],\n",
              "       [0.50178754],\n",
              "       [0.51543206],\n",
              "       [0.49907398],\n",
              "       [0.5072632 ],\n",
              "       [0.50800663],\n",
              "       [0.4993157 ],\n",
              "       [0.49913397],\n",
              "       [0.5147077 ],\n",
              "       [0.49911207],\n",
              "       [0.50911665],\n",
              "       [0.5014527 ],\n",
              "       [0.50571495],\n",
              "       [0.5009626 ],\n",
              "       [0.51506823],\n",
              "       [0.5005568 ],\n",
              "       [0.5140597 ],\n",
              "       [0.51044065],\n",
              "       [0.50912297],\n",
              "       [0.4991866 ],\n",
              "       [0.5023585 ],\n",
              "       [0.50796366],\n",
              "       [0.4991291 ],\n",
              "       [0.50100696],\n",
              "       [0.5005771 ],\n",
              "       [0.4993067 ],\n",
              "       [0.5079311 ],\n",
              "       [0.4991329 ],\n",
              "       [0.5001288 ],\n",
              "       [0.49943545],\n",
              "       [0.50068384],\n",
              "       [0.52261525],\n",
              "       [0.5401607 ],\n",
              "       [0.49912786],\n",
              "       [0.50001043],\n",
              "       [0.50299835],\n",
              "       [0.5027621 ],\n",
              "       [0.50289303],\n",
              "       [0.50024396],\n",
              "       [0.5089869 ],\n",
              "       [0.5052942 ],\n",
              "       [0.49907678],\n",
              "       [0.499074  ],\n",
              "       [0.5012116 ],\n",
              "       [0.5000875 ],\n",
              "       [0.5253622 ],\n",
              "       [0.5004857 ],\n",
              "       [0.49908075],\n",
              "       [0.49975404],\n",
              "       [0.49942532],\n",
              "       [0.50543094],\n",
              "       [0.5123886 ],\n",
              "       [0.5048808 ],\n",
              "       [0.5004773 ],\n",
              "       [0.50039375],\n",
              "       [0.49940926],\n",
              "       [0.50489944],\n",
              "       [0.50283974],\n",
              "       [0.49907663],\n",
              "       [0.51946443],\n",
              "       [0.52377933],\n",
              "       [0.49987382],\n",
              "       [0.5170104 ],\n",
              "       [0.5019523 ],\n",
              "       [0.5005759 ],\n",
              "       [0.5018607 ],\n",
              "       [0.5049467 ],\n",
              "       [0.501991  ],\n",
              "       [0.50156534],\n",
              "       [0.5024522 ],\n",
              "       [0.5019116 ],\n",
              "       [0.51641095],\n",
              "       [0.51310813],\n",
              "       [0.51125777],\n",
              "       [0.50975806],\n",
              "       [0.5028037 ],\n",
              "       [0.49957985],\n",
              "       [0.5084721 ],\n",
              "       [0.50338733],\n",
              "       [0.5148233 ],\n",
              "       [0.51644343],\n",
              "       [0.5009347 ],\n",
              "       [0.5028638 ],\n",
              "       [0.5106009 ],\n",
              "       [0.51686525],\n",
              "       [0.5157829 ],\n",
              "       [0.49907592],\n",
              "       [0.49912393],\n",
              "       [0.5018763 ],\n",
              "       [0.5063924 ],\n",
              "       [0.50815594],\n",
              "       [0.5066221 ],\n",
              "       [0.5000476 ],\n",
              "       [0.5022248 ],\n",
              "       [0.5091725 ],\n",
              "       [0.4990743 ],\n",
              "       [0.5106847 ],\n",
              "       [0.50452375],\n",
              "       [0.5108395 ],\n",
              "       [0.5195113 ],\n",
              "       [0.5062886 ],\n",
              "       [0.49952754],\n",
              "       [0.50262845],\n",
              "       [0.4991365 ],\n",
              "       [0.50960404],\n",
              "       [0.52867967],\n",
              "       [0.4992268 ],\n",
              "       [0.51258296],\n",
              "       [0.49892846],\n",
              "       [0.4990758 ],\n",
              "       [0.50763243],\n",
              "       [0.49944645],\n",
              "       [0.51021475],\n",
              "       [0.5097107 ],\n",
              "       [0.50839883],\n",
              "       [0.49910274],\n",
              "       [0.5023984 ],\n",
              "       [0.5212775 ],\n",
              "       [0.5161505 ],\n",
              "       [0.49909008],\n",
              "       [0.50783426],\n",
              "       [0.5035395 ],\n",
              "       [0.51014495],\n",
              "       [0.5074391 ],\n",
              "       [0.49907404],\n",
              "       [0.5056352 ],\n",
              "       [0.5205036 ],\n",
              "       [0.49908745],\n",
              "       [0.4990797 ],\n",
              "       [0.51431704],\n",
              "       [0.50708824],\n",
              "       [0.5054015 ],\n",
              "       [0.49928078],\n",
              "       [0.49955902],\n",
              "       [0.49907753],\n",
              "       [0.5024159 ],\n",
              "       [0.5115447 ],\n",
              "       [0.51244056],\n",
              "       [0.5086466 ],\n",
              "       [0.5165363 ],\n",
              "       [0.49907398],\n",
              "       [0.49933764],\n",
              "       [0.50242543],\n",
              "       [0.5206707 ],\n",
              "       [0.49950776],\n",
              "       [0.504381  ],\n",
              "       [0.5068983 ],\n",
              "       [0.50257134],\n",
              "       [0.50308204],\n",
              "       [0.50827736],\n",
              "       [0.5010532 ],\n",
              "       [0.5065631 ],\n",
              "       [0.5232347 ],\n",
              "       [0.49981654],\n",
              "       [0.49912095],\n",
              "       [0.49908763],\n",
              "       [0.50690186],\n",
              "       [0.50929147],\n",
              "       [0.49907452],\n",
              "       [0.4991225 ],\n",
              "       [0.499408  ],\n",
              "       [0.5015463 ],\n",
              "       [0.50326407],\n",
              "       [0.5162828 ],\n",
              "       [0.51132506],\n",
              "       [0.5142194 ],\n",
              "       [0.52478504],\n",
              "       [0.49943605],\n",
              "       [0.4991653 ],\n",
              "       [0.49907455],\n",
              "       [0.50015765],\n",
              "       [0.5031869 ],\n",
              "       [0.5021794 ]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 247
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "*Here is the graph showing the performance of the training and testing data after fitting the model on the dataset after first try*"
      ],
      "metadata": {
        "id": "E1BzdFMeou4w"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(eval_result,label=\"test\")\n",
        "plt.plot(eval_result_train,label=\"train\")\n",
        "plt.legend(ncol=2)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 286
        },
        "id": "Xsql8M3-a8u0",
        "outputId": "985361f8-7621-464f-c305-b5c7d0000ac0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f16b3899fd0>"
            ]
          },
          "metadata": {},
          "execution_count": 248
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(predictions.shape)\n",
        "seq_predictions=np.transpose(predictions)[0]  # transformation to get (n,)\n",
        "print(seq_predictions.shape)  # now the shape is (n,)\n",
        "# Applying transformation to get binary values predictions with 0.5 as thresold\n",
        "seq_predictions = list(map(lambda x: 0 if x< val else 1, seq_predictions))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mAzYnui_G-Od",
        "outputId": "a18feca3-2d01-4101-e74b-2b9632f3a9fd"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(468, 1)\n",
            "(468,)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "for i in seq_predictions:\n",
        "  print(i)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "T9aSQDXWiAMY",
        "outputId": "1a8f2186-d1fb-441d-bb7f-eb2e47971ce5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "1\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "1\n",
            "1\n",
            "1\n",
            "1\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n",
            "0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "loss,acc,pre,rec=hypermodel.evaluate(x_train_resampled,y_train_resampled)\n",
        "print(acc)\n",
        "loss,acc,pre,rec=hypermodel.evaluate(x_test,y_test)\n",
        "print(acc)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "l3agUFonJtw4",
        "outputId": "28671beb-4381-41c6-eeed-9ceb499f1161"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "54/54 [==============================] - 0s 6ms/step - loss: 0.6931 - accuracy: 0.8786 - precision_2: 0.8278 - recall_2: 0.9895\n",
            "0.878609299659729\n",
            "15/15 [==============================] - 0s 6ms/step - loss: 0.6931 - accuracy: 0.6047 - precision_2: 0.4290 - recall_2: 0.9433\n",
            "0.6047008633613586\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# save the file in json\n",
        "model_json = hypermodel.to_json()\n",
        "with open(\"model_acctest_69.json\", \"w\") as json_file:\n",
        "    json_file.write(model_json)\n",
        "# serialize weights to HDF5\n",
        "hypermodel.save_weights(\"model_acctest_69.h5\")\n",
        "print(\"Saved model to disk\")"
      ],
      "metadata": {
        "id": "O35nbLPnAQ5c"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# load the model from json\n",
        "json_file = open('model_acctest_69.json', 'r')\n",
        "loaded_model_json = json_file.read()\n",
        "json_file.close()\n",
        "loaded_model = model_from_json(loaded_model_json)\n",
        "# load weights into new model\n",
        "loaded_model.load_weights(\"model_acctest_69.h5\")\n",
        "print(\"Loaded model from disk\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "l8oCjaEfBBaj",
        "outputId": "39e9df92-fb1c-498e-bdc5-47c2d4e8cf99"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Loaded model from disk\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "***The saved and loaded model is compiled and evaluated again on the training and test dataset to give us the following results***"
      ],
      "metadata": {
        "id": "1x_9s1aTpFE0"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# loading the model from disk and calculating the recall and accuracy\n",
        "loaded_model.compile(loss='binary_crossentropy', optimizer='SGD', metrics=['accuracy',keras.metrics.Recall(),keras.metrics.AUC()])\n",
        "score = loaded_model.evaluate(x_test, y_test, verbose=0)\n",
        "score_train=loaded_model.evaluate(x_train_resampled,y_train_resampled,verbose=0)\n",
        "print(\"For testing data\")\n",
        "print(\"%s: %.2f%%\" % (loaded_model.metrics_names[1], score[1]*100))\n",
        "print(\"%s: %.2f%%\" % (loaded_model.metrics_names[2], score[2]*100))\n",
        "print(\"%s: %.2f%%\" % (loaded_model.metrics_names[3], score[3]*100))\n",
        "print(\"For training data\")\n",
        "print(\"%s: %.2f%%\" % (loaded_model.metrics_names[1], score_train[1]*100))\n",
        "print(\"%s: %.2f%%\" % (loaded_model.metrics_names[2], score_train[2]*100))\n",
        "print(\"%s: %.2f%%\" % (loaded_model.metrics_names[3], score_train[3]*100))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ioze9Dd7Bk3L",
        "outputId": "efc6bbf6-0795-41e9-db04-3cfe97ac1b1c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "For testing data\n",
            "accuracy: 63.25%\n",
            "recall_1: 88.73%\n",
            "auc_1: 73.35%\n",
            "For training data\n",
            "accuracy: 90.15%\n",
            "recall_1: 95.92%\n",
            "auc_1: 90.20%\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(score,label=\"test\")\n",
        "plt.plot(score_train,label=\"train\")\n",
        "plt.legend(ncol=2)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 286
        },
        "id": "y-IBV_j35T1T",
        "outputId": "29fd2481-895b-4e62-b6a7-36dddf1e0d6a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f16b28c1a50>"
            ]
          },
          "metadata": {},
          "execution_count": 263
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXgAAAD7CAYAAABgzo9kAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3deXhU1f3H8fdsyWTf2GSTTQ5LWIMgq0DrrlXRVnDB2loXKFqX+nOrS1u1Lq2tBYtLawUUXAtYra0CIjsS1ggeUJB9zWRPZjKZub8/MkAIgUzCTO7M5Pt6Hp5Mbu6d+R5u8rl3zr1zjsUwDIQQQsQeq9kFCCGECA8JeCGEiFES8EIIEaMk4IUQIkZJwAshRIyym10AQG5ubjxwLrAf8JlcjhBCRAsbcBbwVU5Ojqf2DyMi4KkO9yVmFyGEEFFqJLC09sJICfj9AN27dycuLq7BG+fl5ZGdnR3yoswgbYk8sdIOkLZEqsa2pbKykq1bt0IgQ2uLlID3AcTFxREfH9+oJ2jsdpFI2hJ5YqUdIG2JVGfYljq7tuUiqxBCxCgJeCGEiFES8EIIEaMk4IUQIkZJwAshRIySgBeiOZHhwZuVSLlNUghxhvyecqqK86kqyaeq+Ai+YlfgcfUyX0k+6e4ydixKwBrnxBrnxOIIfI1zYo1LOP7Y4cQSV2O9uFNvY3HEY7FYzG6+qIMEvBARzjAM/O4yfEfDuvhIILhd1csCy43KipO2tSWlY0/NwpHRhoSze3PQVUSbFpn4K934vW6MSjf+ygr85cVUFR7CX+nG8LrxeyrA8AdZoQVLXHzgoFDrQBHnxOKoeaBwBh4n1DiQOGscPBKwxjsDBw3pYDhTEvBCmMgwDPwVJceC+1iIl1SfffsCZ9+Gt9YwIxYrtuR07ClZxLVoT0LnfthTs7CnZGFPzcKWkoU9JQOLzXHCZjtzc8nKyQmqLnxV1eFfGTgQeKsPBtUHBTdG4GfHDgpH1zt20CihquhwjeUV4A9+qKkT3inUcfBIKCzGVbylxnq1DiSOhFoHlOZ30JCAFyJMDMOPr6w4ENpHanSVuI4/Ls7H8HlP3NBixZ6SiS01i7jWnUg8J6c6sGsGeHIGFqstbLVbLBawO7DZHdgSU0P2vIbPi9/jxu+teaBwH39H4akIHEiOLzcCBxZ/pRu/u5Sq4sMYlW7iKsoo3LsBfFXBt8sRH0TXVO13IQknvNM4up41PtA9Fcb9cKYk4IVoBMPvw1dWRFVx/vEAD5x9Hw9wF/hrhY/Vjj01E3tKFvFtu2FXQ44F99EQtyWlRXRonAmLzYEt0YGNlDN+rtzcXHJycqoPGpWe07yjqPkupOLkg4e7rPpdUo0DzkkH3dO1yR53im6mYA4eTmwJZ/5/cSoS8ELUYvh9+EoLjnWbxO/YQL5r47F+76Nn4bX7qC02B7bU6pB2duhRo6vkeLeJLSnVlG4CwzCY8z/NkrWHSG7hQp2d2eQ1hIvF5sCW4ICE5JA9p+Grwu+tddA4eqDwHu+GOuFdSM2DiqcCf0nBseVGpRujqvKUr2cfNB6ov+usoSTgRbNi+LxUlRSc1G1S88zbV1Z4QngnAsWO+GNBndApu/pxSlZ1oAeWWxNSIvJuEsMweG1eHh8t2Y7DbuGBvy7hR6O6csPFPXDGSQTUxWKzY7PZwZkUsuc0/L4TrlEcDX8Mg4Ij7pC9Tk2yd0XM8FdVHu8eOeGC5fE7TnxlhSdtZ4lzYk9tgT01i8SWHU7q787bvpsBQ4ZHZHjXx+c3mPbeej5bvYsfjepCz1ZuNuyNY+7i71iVd4ApP+lPn24tzC6zWbBYbVicSVjrOmjk54blNSXgRVTwV7qpKnGdfMGyxh0n/vLik7azOpOOdZPEt+l8vNuk5pl3fOJpX9vYfSQqw73K5+fF2Wv5ct1errugOzdc1IO1a9cy+dp+jOzflqnvbuDhvy3j4qGduOXyXiQ6HfU/qYgqEvDCdH5PxfEP5NQ68/YFzr797tKTtrMmpBwL6fh25xx7fLzrJBNrXIIJLTJfpdfHczPXsOrrA9x8WS+uHXvOCT/v260lL90/mrc+/Yb5X37Hms0HmPzj/gzq2dqkikU4BBXwSqnuwJtAFpAPTNRab6u1ThvgFaAz4ACe0lrPCm25IqoYBj532bF7uY+eedf8cE5ViQvDU37SpraktOoz7bTWODv0OunM25aSidURO5M9hJK7soqn3ljN+q2Huf3qPlw+okud6znj7Pz8R9mM6NeWv7yznidfX8nonPb84so+pCY1fGY1EXmCPYOfDkzTWs9SSt1IdZCPrbXOn4A1WusrlVItgVyl1GKt9e4Q1iuigFHl5cB7z5K+M4+d/619u5nl2Ad0HFntSOjUtzq0awZ4ciYWu3QXNEa528tv/76KLTvyufu6/vxw8Nn1bqPOzuQv957PO59v5f0F21inD3H71X0Z0a9tVHZNiePqDXilVCtgIHBBYNFsYKpSqqXW+nCNVfsBLwJorQ8rpdYDPwH+GNqSRaQryv2Uiu3rqGzfn7NUv2MBbk8JfEDHJj2D4VBSXsnjr65g+94i7r9hECMHtAt6W4fdxo0X92R437a89M46npu5hi/XteHOa/qRmeoMY9UinIK5IbcDsFdr7QMIfN0XWF5TLjBeKWVRSnUGhgH1nz6ImOJzl1G47H0SuvSjPPtS0s/7Ecm9huNs3wN7WksJ9zApKHHz8MvL+H5/MQ//dHCDwr2mzm3TeOGuUfz0sl7kfnOISc8t5PPVO6uHLhBRx1LfjlNK5QAztNa9ayzbDNyotV5bY1lLqs/g+wC7gHKqDwz31ldEbm5uJ2BHYxogIkuCXkT8jhWUDPs5vlS5YNcUisqrmLHgCMXlPsafn0XXNqE54z5S7GX+qgJ2Ha6ka5t4Lh+cQUayHKAjVOecnJzvay8MZm/tBtoppWxaa59Syga0DSw/JtBdc+PR75VSnwCbG1JhdnZ2o2YWP/qR5VgQzW2pKj7C7s9zScoeRdcxl0Z1W2qK5HbsP1LGy68sp8ILv7tjOL27ZJ12/Ya25YLzDf6zfAf//Hgzr3x6mJsv68WlwzpjtZrfNx/J+6WhGtsWj8dDXl7eKX9ebxeN1voQsB6YEFg0AVhXq/8dpVSWUsoeeDyW6jP5txtcsYharsXvYBh+MkZPqH9lccZ2HyzhwWlLqXB7eSqIcG8Mq9XCZSO6MO3XY+nZKZNX/rWJh15eyp5DJSF/LRF6wQ6KcQcwRSm1FZgS+B6l1CdKqUGBdQYDW5RS3wC/Ba7QWp98/5uISZWHdlK6cRFpgy7FkdbK7HJi3va9RTw4bSl+w+CZSSPo1iE9rK/XKjORJ28byt3XDWDngRLu+uMXvL9wGz5fsGPGCzME1aGmtf4GGFLH8ktrPP4PcE7tdUTzkL9wFlZnIunDx5ldSsz7ZqeLJ15bSUK8nafuGEbblqEbZOt0LBYLPxzckYE9WjH9w428+fFmlm7Yy93XDaBz27QmqUE0TPMa/V6ERcXOPCq+W0v6sHFhHfpUwKZvj/DYK8tJTYzj2ckjmizca8pMdfLQzefyfxMHkV/o5p4XFzPrP1vwVgU/mYdoGnJJXJwRwzBwLZiJLbUFqYMuMbucmLZmy0Ge+edqWmcl8fs7hpl6f7rFYmFEv3b07daS1+dt4p3Pt7J8037uuq4/PWJoKOJoJ2fw4oyUbVmOZ/+3ZJ4/XoYOCKPlG/fx1BuraN86hWcmDY+YDx+lJsVx7/U5PH7reVS4vTzw1yW8Pi8Ptyf4WZZE+EjAi0YzfF5ci94irlVHkrNHmV1OzFqUu5tnZ66hW/t0nrpzOGnJkXcgHdSzNdMeGMvF53Vi3pffMeWPi9j47eH6NxRhJQEvGq147WdUFR4kc8xNMTvFnNn+s+J7Xpy9luwuWfz29mEkJ0TuGD2JTgeTru3H03cOx4KFR/62nKnvraesIvjp70RoScCLRvF7yilY+h7Os7NJ6DrA7HJi0tzF3/Hy+xvI6dGax249j4T46Lhk1qdbC166fzRXnd+Vz1btZPLzC1m9+YDZZTVLEvCiUQpXzMNfXkzW2JtkxMEQMwyDOZ9p/j4/j+F92/LwTwcT74iud0hHhyJ+/q5RJCU4+N3fV/HCrFyKSj1ml9asSMCLBqsqcVG0aj5JvYYT37ab2eXEFMMwePPjzbz16TeMHdSBX9+Yg8MevX+m3Ttm8Od7RjPhQsXSDXuZ/PxClqzbK4OXNZHo/c0Rpin48h0Mv5/M0debXUpM8fsNXvnXJj5Y9C2XDOvE3dcNwGaL/j9Rh93K9Rf14M/3jqZlRiLPzVrDU2+sJr+owuzSYl70//aIJlV5ZA8lGxaSmnMhjow2ZpcTM3w+Py+9u46Pl+3g6tHduHNc34gY0CuUOp2VygtTRnLL5b1Ypw8x+bmFfLZKhiIOJwl40SCuRbOwxDnJGH6t2aXEDG+Vn+ffymXBV7u5/qIe3HJ5r5i9rmGzWRk35hxeun8Mndqm8dK763ns1RUcdMmwVeEgAS+C5t69hfKtX5E+9CpsSTL2SChUen088+Zqlm3Yx8+u6M2EC1XMhntN7Vom8/Sdw7nzmr7onS5++fxC5i/5Dr9fzuZDSQJeBMUwDPIXzMCWnEna4MvNLicmVHiqePL1lXy1+SCTrunL1aOb1wVrq9XCpcM6M/XXY+nVJYvX5ubx4LSl7D4oQxGHigS8CEq5XoVn71YyRl0nQxKEQFmFl8dfXUHed0e4Z8IALhnW2eySTNMqI5Enbj2PeyYMYPfBEu7+0xe8t2ArVTIU8RmLjk9OCFMZvipci97C0aI9Kf3GmF1O1Csq9fD4ayvYub+YByaey/C+bc0uyXQWi4WxgzoyoHsrpv9rIzM+2cLSDfu4+7oBdGkn3YGNJWfwol4l6z/H69pH5pgbZUiCM+QqdvPw35ax+0AJj9wyRMK9loxUJw/dPJgHbz4XV7Gbe/+8mJkyFHGjyRm8OC1/ZQUFS97D2aEniecMqn8DcUqHCsp5dPpyCordPP6L8+jbraXZJUWs4X3b0rdbC16fl8e7n29lxaZ93PWTAfToJEMRN4ScwYvTKlr5Eb6yQjJ/MLFZ3N0RLvsOl/LgtKUUl3r43e3DJNyDkJIYxz0TBlYPRezx8cDUJbw2d5MMRdwAcgYvTqmqtJDClfNI6jEUZ7vuZpcTtXYeKOY305fj8xs8dedwurYP7/ypsWZQz9ZM+/UY3vx4M/OXbGfV1weY8uP+ZpcVFeQMXpxS4ZJ3MXxeMsfIkASN9e3uQh6atgyLBZ6ZJOHeWIlOB3de049nJg3HarXw6CvLmb+qQIYirocEvKhTZf4+itd9RuqAC3BkyoXAxti8I59Hpi8jId7GHyaPpGObVLNLinrZXVvw1/vHMG50N9ZtL2PScwtZ/bUMRXwqEvCiTgVfvIXFEUf6iB+bXUpU2rD1MI+9uoL05Hj+MHkkZ7VIMrukmBHvsHHLFb259cJWpCQ6+N0/VvH8rDUyFHEdJODFSdx7t1L2zUrSz7sSe7J0KTTU6s0HePLvK2mTmcgfJo+gZUaC2SXFpHZZcbx4z2iuv1CxfOM+Jj23kC/X7ZHBy2oI6iKrUqo78CaQBeQDE7XW22qt0wp4A+gAOIBFwF1aa7nkHUUMw8C1YAa2pHTShlxhdjlRZ+mGvbwwK5fO7dJ48hdDSU2KM7ukmOawW5lwUQ+G9W3LX95Zx/Ozcvly3V7uvKYvWWlyYA32DH46ME1r3R2YBrxSxzoPA1u01n2BvkAOMC4kVYomU75tDe7dW8gY+ROscfIH0hALvtrF8zPXoM7O4Pe3D5Nwb0Jnn5XK81NGcsvlvY8NRfzflTIUcb0BHzgzHwjMDiyaDQxUStW+kdcAUpRSViAeiAP2hrBWEWaG34dr0SwcmW1J6f8Ds8uJKh8v28Gf56yjb7eWPPmLoSRF8OTYsap6KOJu/DUwFPHU99bzm1eWcyC/zOzSTBPMGXwHYK/W2gcQ+LovsLym3wHdgf3AAeC/WutlIaxVhFnJhkV4j+whc8wNWGzyEYlgfbhoG9M/3MiQ3m34zc+H4IySybFjVdvAUMSTrunL1l2F/PKFRcz/8jt8zXAoYkt9b2GUUjnADK117xrLNgM3aq3X1lh2O9ALuAdIAf4D/Elr/X59ReTm5nYCdjSmASJEqipJWzIdf0IaJUMmgnxqtV6GYfDFpmIW55XQu2MC44ZlYouxWZiiXVFZFR+tLuTb/W7at4jjyiEZtEyLyXdXnXNycr6vvTCYU43dQDullE1r7VNK2YC2geU1TQF+prX2A0VKqXnAGKDegD8qOzub+PiGD0Wbm5tLTk5Og7eLRGa1pWDp+xR4Sml/3UM4O/QIyXPGyn6pqx2GYfCPj75mcd5eLhjckck/7h8V4R4r+wSCb8uYkQaLcvfw+rxNvPLpYSZcqBg3phv2CJrvtrH7xePxkJeXd8qf19tCrfUhYD0wIbBoArBOa3241qo7gIsBlFJxwA+BU7+yiBi+siIKV8wlsfu5IQv3WOb3G7z8wUbmLv6Oy0d05pdREu7NVfVQxB2Y9sBYhvRuw8z/bOG+P3/Jd3sKzS4t7II9hN0BTFFKbaX6TP0OAKXUJ0qpo0MM/goYqZTaRPUBYSvwWojrFWFQsOx9DK+HzDE3ml1KxPP5/Px5zlo+XfE91449h9uu6hNzk2PHqowUJw/efC4P3XwuBSVu7v3Ll8z4ZDOV3tgdijioq0Fa62+AIXUsv7TG4++AC0JXmmgK3oIDFOf+j5T+PyCuRXuzy4lo3io/z89aw4pN+7npkp785IcyAFs0GnZ0KOL5eby3YBsrNu3nrp8MoGfn2BuKOHI6oYQpXF+8jcVmI2PkdWaXEtE8Xh9PvbGKFZv2c+uV2RLuUS45MY5fjR/Ik78Yisfr4/+mLeHVuZuoiLGhiCXgmzH3vm8p27yMtMFXYE/JMLuciOXx+nnytZWs1Yf45Y/7ceWormaXJEJkYI9WTL1/DJcO68xHS7bzyxcWsX7rIbPLChkJ+GbKMAxcC2diTUwlfeiVZpcTsUrLK5m58Ahf78jn3utzuOi8TmaXJEIs0engjnF9+cPkEditFn7zygpeemcdpTEwFLEEfDNV8d063DvzyBjxY6zxiWaXE5GKSj088rfl7C+o5MGJ5zJ6oFyjiGW9u2Tx0v1juGZMNxZ8tYvJzy1kVd5+s8s6IxLwzZDh95G/cCb2jDakDpTr4nXJL6rgwWlL2XO4lAmjWjC0z1lmlySaQLzDxk8v780Ld48iNSmO37+xmudmRu9QxBLwzVDppsV4D+8ic/T1WGwx+am+M3LQVc6D05aSX1TBk784j25tnWaXJJrYOR0y+NOvzueGi3uwYtM+7nx2IV+sjb6hiCXgmxm/14Nr8Rziz+pGUs9hZpcTcfYcKuHBqUsoKffy+zuGk921hdklCZM47FbGX6D4872jOatFIn98K5ff/WMV+UUVZpcWNAn4ZqZ4zX/wleST+YObsMh4MyfYsa+Ih6Ytw+vz88yk4XTvKHcWCTi7TSrPTRnFz3/Umw3bjjDpuYX8d+X3UXE2LwHfjPjKSyhc9gGJ3XJIODvb7HIiytZdBTz88jJsNgvPTBpB57ZpZpckIojNauGq87vx1/tH07VdOlPf28Cj0yN/KGIJ+GakcPkH+CvdMiRBLV9vz+fR6ctJSnDwh8kj6NA6xeySRIRq2yKZ398xjEnX9mPb7uqhiOdF8FDEEvDNhLfwEEVr/kNyn9HEtepodjkRY50+xGOvriAz1ckfJo+gTZZMji1Oz2q1cMnQTkz79Vj6dG3B6/Py+L+pS9h1oNjs0k4iAd9MFCyejcViJfP88WaXEjFW5u3nt39fRdsWSfxh8ghapMsUhSJ4LTMSeOznQ7jv+oHsO1zK3X9azDufaap8frNLO0YCvhnwHNhOad6XpA2+DHtqltnlRITFa/fwzJtf0aVdKk9PGk56SsPnIRDCYrEwOqcDLz/wA87LbsOsT7/h3j8v5tsIGYpYAr4ZcC2chTUhmbShV5tdSkT4bNVO/vh2Lj07ZfK724eRkiiTY4szk54Sz/9NPJeHfzqYwhIP9/3lS9782PyhiGXyyBhXvn0DFTs2kPnDn2JzSv/y/CXf8drcPAaqVjz003NxxsmfgAidoX3Ook/XLP4+/2veXxgYivi6/vTqbM47ZzmDj2GG4ce1cCb2tFak5Vxsdjmme2/BVl6bm8fQPmfx6M8GS7iLsEhOjOPu8QN48raheKt8PDhtKa/8a6MpQxFLwMew0rwlVB7cUT0kgb35DklgGAYzPtnMjE+2MHpge/7vpkE47DazyxIxbqBqxdRfj+Wy4Z35eNkOfvn8Qtbpph2KWAI+RvmrKilYPJu4Nl1I6j3c7HJMYxgGr82rnrnnovPO5lcTBmKLoMmWRWxLiLdz+9V9eWbSCBx2G4+9uoK/zFlHaXllk7y+/KbHqOLcT6kqOkzm2BuxWJrnbvb5Daa+t4GPlmznR6O6MPnafjI5tjBF7y5ZvHTfaK4dew4Lc3cz+fmFrNgU/qGIm+dffozzVZRSuPQDErr0I7FzP7PLMUWVz8+f3s7lf6t2ct0Pu3Prj7Jl7B1hqjiHjZsv68Uf7xpFWnI8T/9zNc/O+IrCkvANRSxXmWJQ4Yp/4XeXkTnmJrNLMYW3ysezM9aw6usD3HxZL64de47ZJQlxTLcO6fzpV+fzwcJtzPlsKxu2HebmseGZ8FsCPsZUFR+hePXHJPcZRXybzmaX0+TclVU8/cZq1m09zO1X9+HyEV3MLkmIk9htVq67QDG0z1n864vvsFjC0ycvXTQxxrV4DgYGGc1wSIJyt5cnXlvJhm2Hufu6/hLuIuJ1bJPK3eMHkJUSnnPtoJ5VKdUdeBPIAvKBiVrrbbXWmQH0rbGoL3CV1np+iGoV9ag8tJPSjV+QNuQKHGmtzC6nSZWUV/L4qyvYvreI+28YxMgB7cwuSQjTBXvYmA5M01rPUkrdCLwCjK25gtZ64tHHSql+wELgv6EqVNQvf+FMrM5E0oePM7uUJlVQ4uaxV1aw51ApD/90MIN7tzG7JCEiQr1dNEqpVsBAYHZg0WxgoFKq5Wk2+znwltY6OmeqjUIV32+i4rt1pA8bhy2h+YxnfqSwgoemLWV/fhmP3zpEwl2IGiz1TTullMoBZmite9dYthm4UWu9to7144B9wA+11uuDKSI3N7cTsKMBdYuaDIOUFf/EWllG0cg7wNY8rp27SquYseAwFZV+rj+/BWe3khEhRbPVOScn5/vaC8ORBFcBu4IN95qys7OJj2/4H2lubi45OTkN3i4SNaYtpZuXcah4Py2v+CXd+g4JU2UNF879svtgCS9NX47PsPLM5OGc0yF886c299+vSCVtAY/HQ15e3il/HsxdNLuBdkopG0Dga9vA8rr8DPhHA+sUjWT4vLgWvUVcq7NJzh5ldjlNYvveIh56eSl+w+CZSSPCGu5CRLN6A15rfQhYD0wILJoArNNaH669rlKqPTASeCuURYpTK177P6oKD5I59iYs1tgfQEvvdPHw35bhsNt4dvIIzj4r1eyShIhYwd4HfwcwRSm1FZgS+B6l1CdKqUE11rsZ+EhrXRDaMkVd/J5yCpa+j7NTHxK69De7nLDb9O0RfvPKclIT43h28gjatkw2uyQhIlpQffBa62+Akzp3tdaX1vr+qRDVJYJQuGIu/vJissbeFPPjrOR+c5Cn31hN66wkfnf7ULLSZP5UIerTPG63iEFVxfkUrfqIpN4jiD+rq9nlhNXyjft4ftYaOrZJ5be3DSUtWe6WESIYEvBRqmDJuxh+P5mjrze7lLBalLubP89ZR/cO6Tz+i6EkJzTfiUuEaCgJ+ChUeXg3JRsWkjroEhzprc0uJ2w+XfE9L3+wgT5dW/Doz4aQEC+/rkI0hPzFRCHXorewxDnJGHGt2aWEzdzF3/H3+XkM6tmaB28+l3hH7N8hJESoScBHmYpdmynf9hUZo2/Alhh7twgahsE7n2/lrU+/YXjfttx3Qw4Ouwx6KkRjSMBHEcMwcC2ciS0lk7TBl5ldTsgZhsGbH2/mg0XfMnZQB+76SX+ZP1WIMyABH0XK9Eo8e7fS4rI7sTpi604Sv9/g1bmb+HjZDi4Z2ok7xvXFKvOnCnFGJOCjhOGromDRWzhatCel7xizywkpn9/gr++uY8FXu7l6dDduubxXzN/XL0RTiPqAd3uq2LyrHHvqYTJTnWSmOkl02mMuIErWf47XtZ/WP34wpoYk8FZVT469dMM+rr9QMf5CFXP7TgizRH3Ab/z2CO8udfHu0uXHlsXH2Y6FfVaqk8w057HvM2t8Hy233fk9FRQseRdnx14knjOo/g2iRKXXxx9mfMVXmw/ysyt6c/XobmaXJERMiY6EO43Bvdtw1xVtaNO+K/nFbgqK3biK3biK3OQXu9m2pxDXZjeeSt9J2ybE26sPAoHAz6jjoJCRGo8zztz/psJV8/GVFVWfvcfI2a3bU8Xv31jFhm1HmHRNXy4Z1vwmCBci3KI+4AEyU+z06dbilD83DINyd1V18Nc4ALiKqw8CriI3W7534Sp2463yn7R9UoKDzNT4k94FZKUm1HhHEI/DHvquk6rSAopWziepx1Cc7bqH/PnNUFbh5cnXV6J3urhnwgDGDupodklCxKSYCPj6WCwWkhIcJCU46ND61NPZGYZBWYX3WOi7ar0bKCh28/X2fFzFbqp8J8+ElZLoOKkbKOvoO4Oj7whSnA26r7twyXsYPi+ZY2JjSIKiUg+Pv7aCnfuLeeCmcxner63ZJQkRs5pFwAfLYrGQnBhHcmIcZ7c59YeI/H6DkvLKE98NlJx4UNh9sARXiQe//+QDQVpy3InvBmocAA7lV9KpqIL05Hh8hfspXvcZqQMvxJEZ/UHoKnbzm1eWs/9IGY/cMoRBPWN3mAUhIoEEfCNYrRbSkuNJS46nc9u0U67n9xsUlXkoKPZUdwcV1e4iqmDHviIKSzzUPA689t//YbHAbWlf0s1q4++7upD03vqTDgaZqU7SkuOxRcH94ocKynl0+nIKit088Yvz6NvtdHO2CyFCQQI+jKxWCxkp1a2De5YAABbXSURBVN0yXdqd+kDg8xsUlXpwFblZvTaPjFbtqdy7lV76e9YnDWdfqQ3XvgMUlXqoPUe61QLpKUevCRw/AGSkHL94nJnqJDUpzrQPDu07Usqj05dTXuHld7cPo0enTFPqEKK5kYCPADar5VgQFx1KYODAs9k/83W8SelcNelOxsVVT25R5fNTWHL83UBByYkXiw+6ytnyvYvisso6XyPj2DWB+BoXip1kpiYce1eQkugI6Z06Ow8U85vpy/H5DZ66czhd26eH7LmFEKcnAR+Byrd+hXv3FlpccjvWuOMzF9ltVlqkJ9Ai/fSzGXmr/NXhX+ti8dEuon1Hyvh6ez4l5d6TtrXbrMfvGKrxDqDmraRZqU6SEuo/EOxzVfLHuctw2C08M2k4HU9zXUMIEXoS8JHG78e1aBaOrLak9P9Bo57CYbfSKiORVhmJp13P4/Ud/9xArVtHCwIXijdsPUyZu+qkbePs1hM/QJbmJDPl+EHBU+njzQWHSUt28vs7hnNWi6RGtUUI0XgS8BEmbu8GvPl7aX3NA2EfkiDeYaNNVhJtsk4fvm5PVa27hDwnHBB27CtizZaDuGt9mCwzxc4zk0fUe6ARQoSHBHwE8Ve6Sfh2CfHtFYlqsNnlHOOMt9M2Ppm2LZJPu16520tBSfXF4qIyD0bZXgl3IUwkAR9Bilb/G6unlKyxE6NySIJEp4NEp4N2LasPBLm5B0yuSIjmLaiAV0p1B94EsoB8YKLWelsd6/0E+A1gAQzgh1rrg6ErN3b5yoooXDGXylbdcXboYXY5QogYEOxn5qcD07TW3YFpwCu1V1BKDQKeAC7QWmcDI4CiENUZ8wqWvo/h9VDRfbTZpQghYkS9Aa+UagUMBGYHFs0GBiqlan8U8R7gBa31AQCtdZHW2h3KYmOV17Wf4rX/JaX/D/Ann3rQNCGEaIhgzuA7AHu11j6AwNd9geU19QK6KKW+VEqtVUo9qpSKvo5kE7i+eBuLzU7GyOvMLkUIEUNCeZHVBvQFLgDigE+BXcCMYJ8gLy+v0S+em5vb6G3NZCvaR+qW5VR0Hc6GrduB6G1LXWKlLbHSDpC2RKpwtCWYgN8NtFNK2bTWPqWUDWgbWF7TLuB9rbUH8Cil5gGDaUDAZ2dnEx/f8Mmkc3NzycnJafB2ZjMMg/1vzaMyMZWeV9+JNT4hattSl1hpS6y0A6QtkaqxbfF4PKc9Ma63i0ZrfQhYD0wILJoArNNaH6616tvAhUopi1LKAfwA2NDgipuRiu/W4t75NRkjfow1/vTDDwghREMFexfNHcAUpdRWYErge5RSnwTungGYAxwCNlN9QPga+Htoy40dht9H/sJZ2DPakDrwArPLEULEoKD64LXW3wBD6lh+aY3HfuDewD9Rj9JNi/Ee3kWrcfdhsTnMLkcIEYOCnztOhIzf68G1eA7xbc8hqcdQs8sRQsQoCXgTFH/1Cb6SfDLH3hSVQxIIIaKDBHwT85WXULj8QxK75ZBwdm+zyxFCxDAJ+CZWuOx9/JVuMsfcaHYpQogYJwHfhLyFhyjK/ZSUvqOJa9XR7HKEEDFOAr4JFSyejcViJWPUeLNLEUI0AxLwTcRzYDuleV+SNvgy7KlZZpcjhGgGJOCbiGvhTKwJKaQPvdrsUoQQzYQEfBMo376eih0byRhxLVanTD4thGgaEvBhZhh+XAtnYU9rRerAi8wuRwjRjEjAh1lp3hIqD+4gc/T1WOwyJIEQoulIwIeRv6qSgi/eJq5NF5J6Dze7HCFEMyMBH0bFaz6lqvgIWWNvwmKR/2ohRNOS1AkTX0Uphcs+IKFLfxI69zW7HCFEMyQBHyaFK/6F311G5tibzC5FCNFMScCHQVXRYYpXf0xyn/OJb93J7HKEEM2UBHwYuL6cA0Dm+TIkgRDCPBLwIeY5+D2lGxeTeu4l2NNaml2OEKIZk4APMdeiWVidiaQPG2d2KUKIZk4CPoQqvt9ExXfrSB9+DbaEFLPLEUI0cxLwIWIYfvIXzMSe2oLUQZeYXY4QQkjAh0rZ5uVUHviOjPMnYLXHmV2OEEJIwIeCUeXF9cVbxLU6m+TskWaXI4QQANiDWUkp1R14E8gC8oGJWutttdZ5ApgE7AssWqa1nhy6UiNX8dr/UlV4iDbjH8VitZldjhBCAEEGPDAdmKa1nqWUuhF4BRhbx3oztNb3h6y6KOB3l1Gw9H0SOvUhoUt/s8sRQohj6u2iUUq1AgYCswOLZgMDlVJykzdQuGIu/ooSMsfehMViMbscIYQ4xmIYxmlXUErlUH1m3rvGss3AjVrrtTWWPQHcCriAA8DjWusVwRSRm5vbCdjR0OLNZnGXkPbl3/C27k5Zv6vMLkcI0Xx1zsnJ+b72wmC7aIIxHXhKa+1VSl0AzFNK9dRa5wf7BNnZ2cTHxzf4hXNzc8nJyWnwdmfq8L9fpgToOu6XONJbh+Q5zWpLOMRKW2KlHSBtiVSNbYvH4yEvL++UPw/mLprdQDullA0g8LVtYPkxWusDWmtv4PFngZ9nN7jiKFF5eDclGxeROujikIW7EEKEUr0Br7U+BKwHJgQWTQDWaa0P11xPKdWuxuP+QCdAh6zSCONaNAtLnJOM4deYXYoQQtQp2C6aO4A3lVKPAQXARACl1CfAY1rrNcDTgf56H1AJ3KS1PhCGmk1XsetryretIWP0DdgSU80uRwgh6hRUwGutvwGG1LH80hqPbw5hXRHLMAxcC2ZiS8kkbfBlZpcjhBCnJJ9kbaCyb1bi2beNjFHjsToafkFYCCGaigR8Axi+Kgq+eAtHyw6k9B1tdjlCCHFaEvANULzuc7yu/WSOuVGGJBBCRDwJ+CD5PRUULn0XZ8deJHaLjXtvhRCxTQI+SIUr5+ErKyJz7EQZkkAIERUk4INQVVpA0aqPSOo5FGe7c8wuRwghgiIBH4SCJe9i+Lxkjr7B7FKEECJoEvD1qMzfS8m6z0kdcAGOzLPMLkcIIYImAV8P16K3sDjiyBj5E7NLEUKIBpGAPw33nm8o16tIP+8qbElpZpcjhBANIgF/CoZhkL9gJrakdNKGXGF2OUII0WAS8KdQvvUrPHu+IWPUdVjjnGaXI4QQDSYBXwfD78O1aBaOrLak9P+B2eUIIUSjSMDXoWT9Arz5e2VIAiFEVJOAr8Vf6abgy3eIb69I7D7Y7HKEEKLRJOBrKVr9b3xlhWT9QIYkEEJENwn4GnxlRRSu+BeJ3QfjbN/D7HKEEOKMSMDXULD0PQxvJZljZEgCIUT0k4AP8Lr2U7z2f6T0/yFxLdqbXY4QQpwxCfgA1xdvY7HZZUgCIUTMkIAH3Hu3UbZlOWlDrsCekmF2OUIIERLNPuANw8C1cCbWxFTSz7vK7HKEECJk7MGspJTqDrwJZAH5wESt9bZTrKuAdcDLWuv7Q1VouFR8uxb3rq/JuuhWrPEJZpcjhBAhE+wZ/HRgmta6OzANeKWulZRStsDP5oamvPAy/D7yF83EntGG1AEXmF2OEEKEVL0Br5RqBQwEZgcWzQYGKqVa1rH6g8C/ga0hqzCMSjZ+gffwbjLH3IDFFtSbGSGEiBrBnMF3APZqrX0Aga/7AsuPUUr1Ay4CXgx1keHg93oo+HIO8W3PIanHULPLEUKIkAvJaatSygG8CtyitfZVd8M3XF5eXqNryM3NbdD68duXk1jiorDnpRxYu7bRrxsODW1LJIuVtsRKO0DaEqnC0ZZgAn430E4pZQuEtw1oG1h+1FlAV+CTQLinAxalVKrW+rZgi8nOziY+Pj746gNyc3PJyckJen1feTG7Fv2ZhG45dLng6ga/Xjg1tC2RLFbaEivtAGlLpGpsWzwez2lPjOsNeK31IaXUemACMCvwdZ3W+nCNdXYBLY5+r5R6AkiO1LtoCpZ9gFHpJnPsjWaXIoQQYRPsXTR3AFOUUluBKYHvUUp9opQaFK7iwsFbeJDiNZ+S0ncMcS07ml2OEEKETVB98Frrb4AhdSy/9BTrP3FmZYVPwRezsVitZIy6zuxShBAirJrVJ1k9+7dT+vUS0gZfjj01y+xyhBAirJpNwFcPSTADa0IK6UNlSAIhROxrNgFfsX09Fd9vImPEtVidSWaXI4QQYdcsAt4w/LgWzsKe3orUgReZXY4QQjSJZhHwpXlfUnnoezJHX4/F7jC7HCGEaBIxH/D+qkoKvphNXJuuJPUabnY5QgjRZGI+4IvX/Ieq4iNkjb0RiyXmmyuEEMfE9BCKvopSCpd9SEKXASR07mt2OUKcxOv1smfPHtxud5O8nt1uZ8uWLU3yWuHWnNridDpp3749DkfDuphjOuALl3+I310mQxKIiLVnzx5SUlLo1KkTFosl7K9XVlZGUlJs3EXWXNpiGAb5+fns2bOHzp07N+h5Y7bPoqroMMVffUJyn/OJb93J7HKEqJPb7SYrK6tJwl1EJ4vFQlZWVqPe5cVswLsWzwEg8/zxJlcixOlJuIv6NPZ3JCYD3nPwe0o3LSb13Euxp9U18ZQQQsS+mAx418JZWJ1JpA8bZ3YpQkSVv/71r1RWVpq2fbRobDs3bdrEfffdF4aK6hZzAV+xYyMV29eRPnwctoRks8sRIqpMnToVr9dr2vbR4lTtrKqqOu12ffr04Y9//GO4yjpJTN1FYxh+8hfOxJ7agtRBl5hdjhBR5cknnwRg/PjxWK1W/va3vzFt2jS01ng8HoYMGcJDDz2EzWZj6tSp/Pvf/yY+Ph6LxcKMGTN48cUXT9h+5syZpKammtmksKj9/9SuXTsyMjLYsWMHZWVlzJs3j/vuu48dO3bg9Xrp2LEjTz/9NGlpaaxatYpnn32WDz/8kD179nDNNdcwfvx4Fi1ahMfj4amnnmLQoNBNsRFTAV+2eRmVB7bT8kdTsNrjzC5HiAZZuGYXn63eFZbnvmBwR8YOOv0EN48//jhvv/02c+bMISkpiUceeYRzzz2Xp556Cr/fz/33388HH3zAhRdeyD//+U+WLl2K0+mktLQUp9N50vbhUrLxC0o2LMTn81Fks4X0uVP6jSWl7+jTrlO7nQ8++CBbtmxh1qxZJCYmAvDII4+QmZkJwIsvvshrr73G/fefPMFdYWEh/fv357bbbmPBggW88MILzJkzJ2TtiZmAN6q8uL54m7hWnUjOHmV2OUJEvYULF7Jx40beeOMNoPqWztatW5OSkkLHjh154IEHGDFiBKNHjyY5uXl3h1588cXHwh1g3rx5fPTRR3i9XsrLy+nUqVOd2yUmJjJmzBjKysro378/zz77bEjripmAL177X6oKD9Fm/KMyJIGISmMH1X+W3ZQMw+Dll1+mQ4cOJ/3s3XffZe3ataxcuZJx48bx+uuv06NHjyapK6XvaFL6jo6oDzrVDPc1a9Ywe/Zs5syZQ2ZmJh999BHvvvtundvFxR3vabBarfX24TdUbCSh103B0vdJ6NSHhC79za5GiKiVlJREaWkpAGPHjuXVV1/F5/MB4HK52L17N6WlpbhcLgYPHsxdd91F9+7d2bZt20nbx7LTtbO4uJjk5GTS09OprKzkgw8+aOLqjouJM3jnjpX4K0rIHDtRPjQixBn42c9+xsSJE3E6nUyfPp3p06dz5ZVXYrFYcDgcPPzwwzgcDqZMmYLb7cYwDHr16sWFF1540vaxepEVTmxnu3btTvjZyJEjmT9/PhdddBEZGRkMGjSITZs2mVKnxTAMU164ptzc3E7AjuzsbOLj4xu0bVVxPjunTSKl51BaXfWrsNTXlHJzc8nJyTG7jJCIlbaEsx1btmyhZ8+eYXnuukRSt8aZam5tqet3xePxkJeXB9A5Jyfn+9rbRH0XjXvfVrBYyRg9wexShBAiokR9F01yj6HoMQZd01ubXYoQQkSUoAJeKdUdeBPIAvKBiVrrbbXWuQW4B/ADNuA1rfVLoS33FOwN69YRQojmINgumunANK11d2Aa8Eod63wA9NNa9weGAfcppWSWDSHqEQnXwURka+zvSL0Br5RqBQwEZgcWzQYGKqVOGKZRa12stT5aRSLgAOQ3V4jTcDqd5OfnS8iLUzo64YfT6WzwtsF00XQA9mqtfQBaa59Sal9g+eGaKyqlfgQ8A3QFHtJam3NvkBBRon379uzZs4fDhw/Xv3IIVFZWnvDhmmjWnNpydMq+hgrpRVat9XxgvlKqIzBXKfWJ1loHu33gdp9Gyc3NbfS2kUbaEnlipR1Q/4iH0aS5tKW8vByXy9Xg5wwm4HcD7ZRStsDZuw1oG1heJ631LqXUauByIOiAb8x98BA791uDtCUSxUo7QNoSqRrblhr3wdep3j54rfUhYD1w9EbzCcA6rXXt7pmeNR63AMYA0kUjhBAmCbaL5g7gTaXUY0ABMBFAKfUJ8JjWeg1wm1LqQsALWICpWuv/Bfn8NuCMZoLxeDyN3jbSSFsiT6y0A6QtkaoxbamRmXWOmxwpQxWMAJaYXYcQQkSpkTk5OUtrL4yUT7J+BYwE9gM+k2sRQohoYQPOojpDTxIRZ/BCCCFCL+oHGxNCCFE3CXghhIhREvBCCBGjJOCFECJGScALIUSMkoAXQogYJQEvhBAxKlI+6FSvIGeVsgEvARdTPRb9H7TWrzd1rfUJsi1PAJOAfYFFy7TWk5uyzvoopV4ArgE6AX201ieNehRF+ySYtjxB5O+TLGAm1UN2VwLbgNvrGDsqEXgDyAGqgPu11v9u4nJPqwFt+SfwQ+BIYNF7WuunmrDUoCil5gKdqZ71rhSYorVeX2udkP69RNMZfDCzSt0AdAPOAYYCTyilOjVZhcELpi0AM7TW/QP/IipIAuYCo4Cdp1knWvZJMG2ByN8nBvCc1lpprfsA3wF/qGO9+4FirXU34ArgdaVUchPWGYxg2wLVQXh0v0RcuAfcrLXup7UeALwA/KOOdUL69xIVAR/srFLAdVTPBesPHOXnAj9uukrr14C2RDyt9VKt9SmHjQ6I+H0CQbcl4mmtXVrrL2osWgmcXceq1xE4sQi8e1wDXBL2AhugAW2JClrrohrfplF9Jl9bSP9eoiLgqWNWKarfJneotV5HTjwD21XHOmYLti0A45VSG5VS/1NKDW3KIkMoGvZJQ0TNPlFKWYE7gfl1/Diq9ks9bQG4Vym1SSk1t+bQ5ZFGKfW6UmoX8BRwcx2rhHS/REvAN0fTgc5a677A88C8QJ+kME+07ZO/Ut3XO9XsQkLgdG15BOgW6Mb5EPg00JcdcbTWt2qtOwIPU/07FFbREvDHZpWCYxci6ppVahcnvoXrWMc6ZguqLVrrA1prb+DxZ4GfZzdxraEQDfskKNG0TwIXjc8BrtNa19UVEDX7pb62aK33Hl2utZ4BJAMNn8C0CWmtZwJj6jhBCOl+iYqAD3ZWKeA94BdKKWugT/sq4P2mq7R+DZghq12Nx/2pvrsj6OkPI0jE75NgRcs+UUo9TfXdMVdprU81i8R7wO2B9c8BzgU+bZoKgxdMW2rtl4uoHnJ8b9NUGBylVLJSqkON768AXIF/NYX07yVqhgtWSvWg+tbCDAKzSmmtdc1ZpQJnw1OBCwObPau1ftWcik8tyLa8SfUvto/qW8Qe11p/YlrRdVBKvQSMA9pQfYtavta6d5Tuk2DaEg37pDeQB2wFKgKLd2itr1ZKrQcu1VrvU0olAf8EBlDdnge01vPMqPlUGtCWz4HWVF+0LAZ+rbVeaUrRp6CUag3MA5Ko/v92UX1r6tpw/r1ETcALIYRomKjoohFCCNFwEvBCCBGjJOCFECJGScALIUSMkoAXQogYJQEvhBAxSgJeCCFilAS8EELEqP8HEnLNJquwuqQAAAAASUVORK5CYII=\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "***The loaded model is compiled and fit again to give us better results of accuracy and recall which is 68.4 and 91.1 respectively after the sampling process***"
      ],
      "metadata": {
        "id": "rmXB0B-qpXNV"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# retrain the model\n",
        "loaded_model.fit(x_train_resampled,y_train_resampled,epochs=100,validation_split=0.2,use_multiprocessing=True,callbacks=[custom_early_stopping,reduce_lr])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yj9TfrYSCXa6",
        "outputId": "b12a9a3f-c9a5-45c5-cc25-c7d1b8bcb317"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "43/43 [==============================] - 1s 5ms/step - loss: 0.3238 - accuracy: 0.8923 - recall_1: 0.9626 - auc_1: 0.8964 - val_loss: 0.3307 - val_accuracy: 0.9441 - val_recall_1: 0.9441 - val_auc_1: 0.0000e+00 - lr: 0.0100\n",
            "Epoch 2/100\n",
            "30/43 [===================>..........] - ETA: 0s - loss: 0.3229 - accuracy: 0.8917 - recall_1: 0.9571 - auc_1: 0.8948\n",
            "Epoch 2: ReduceLROnPlateau reducing learning rate to 0.0019999999552965165.\n",
            "43/43 [==============================] - 0s 5ms/step - loss: 0.3184 - accuracy: 0.8931 - recall_1: 0.9642 - auc_1: 0.8987 - val_loss: 0.3528 - val_accuracy: 0.9353 - val_recall_1: 0.9353 - val_auc_1: 0.0000e+00 - lr: 0.0100\n",
            "Epoch 3/100\n",
            "30/43 [===================>..........] - ETA: 0s - loss: 0.3132 - accuracy: 0.8969 - recall_1: 0.9705 - auc_1: 0.8989\n",
            "Epoch 3: ReduceLROnPlateau reducing learning rate to 0.0003999999724328518.\n",
            "43/43 [==============================] - 0s 5ms/step - loss: 0.3141 - accuracy: 0.8968 - recall_1: 0.9675 - auc_1: 0.9012 - val_loss: 0.3377 - val_accuracy: 0.9441 - val_recall_1: 0.9441 - val_auc_1: 0.0000e+00 - lr: 0.0020\n",
            "Epoch 4/100\n",
            "31/43 [====================>.........] - ETA: 0s - loss: 0.3081 - accuracy: 0.9012 - recall_1: 0.9644 - auc_1: 0.9036\n",
            "Epoch 4: ReduceLROnPlateau reducing learning rate to 7.999999215826393e-05.\n",
            "43/43 [==============================] - 0s 4ms/step - loss: 0.3129 - accuracy: 0.8968 - recall_1: 0.9675 - auc_1: 0.8983 - val_loss: 0.3348 - val_accuracy: 0.9441 - val_recall_1: 0.9441 - val_auc_1: 0.0000e+00 - lr: 4.0000e-04\n",
            "Epoch 5/100\n",
            "30/43 [===================>..........] - ETA: 0s - loss: 0.3064 - accuracy: 0.8990 - recall_1: 0.9638 - auc_1: 0.9059\n",
            "Epoch 5: ReduceLROnPlateau reducing learning rate to 1.599999814061448e-05.\n",
            "43/43 [==============================] - 0s 5ms/step - loss: 0.3126 - accuracy: 0.8960 - recall_1: 0.9675 - auc_1: 0.8984 - val_loss: 0.3344 - val_accuracy: 0.9441 - val_recall_1: 0.9441 - val_auc_1: 0.0000e+00 - lr: 8.0000e-05\n",
            "Epoch 6/100\n",
            "43/43 [==============================] - ETA: 0s - loss: 0.3126 - accuracy: 0.8960 - recall_1: 0.9675 - auc_1: 0.8973\n",
            "Epoch 6: ReduceLROnPlateau reducing learning rate to 3.199999628122896e-06.\n",
            "43/43 [==============================] - 0s 5ms/step - loss: 0.3126 - accuracy: 0.8960 - recall_1: 0.9675 - auc_1: 0.8973 - val_loss: 0.3343 - val_accuracy: 0.9441 - val_recall_1: 0.9441 - val_auc_1: 0.0000e+00 - lr: 1.6000e-05\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7f819c7d73d0>"
            ]
          },
          "metadata": {},
          "execution_count": 33
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# save the file in json\n",
        "model_json = loaded_model.to_json()\n",
        "with open(\"model_acctest_69.json\", \"w\") as json_file:\n",
        "    json_file.write(model_json)\n",
        "# serialize weights to HDF5\n",
        "loaded_model.save_weights(\"model_acctest_69.h5\")\n",
        "print(\"Saved model to disk\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kGr69sfggbce",
        "outputId": "26595f93-bd4f-42dc-d7c4-3e90c192aca5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Saved model to disk\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# load the model from json\n",
        "json_file = open('model_acctest_69.json', 'r')\n",
        "loaded_model_json = json_file.read()\n",
        "json_file.close()\n",
        "loaded_model = model_from_json(loaded_model_json)\n",
        "# load weights into new model\n",
        "loaded_model.load_weights(\"model_acctest_69.h5\")\n",
        "print(\"Loaded model from disk\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "u4Bb7IHqgolR",
        "outputId": "47bbcdf2-4276-4bb8-8831-568a4ee00cf9"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Loaded model from disk\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# loading the model from disk and calculating the recall and accuracy\n",
        "loaded_model.compile(loss='binary_crossentropy', optimizer='SGD', metrics=['accuracy',keras.metrics.Recall(),keras.metrics.AUC()])\n",
        "score = loaded_model.evaluate(x_test, y_test, verbose=0)\n",
        "score_train=loaded_model.evaluate(x_train_resampled,y_train_resampled,verbose=0)\n",
        "print(\"For testing data\")\n",
        "print(\"%s: %.2f%%\" % (loaded_model.metrics_names[1], score[1]*100))\n",
        "print(\"%s: %.2f%%\" % (loaded_model.metrics_names[2], score[2]*100))\n",
        "print(\"%s: %.2f%%\" % (loaded_model.metrics_names[3], score[3]*100))\n",
        "print(\"For training data\")\n",
        "print(\"%s: %.2f%%\" % (loaded_model.metrics_names[1], score_train[1]*100))\n",
        "print(\"%s: %.2f%%\" % (loaded_model.metrics_names[2], score_train[2]*100))\n",
        "print(\"%s: %.2f%%\" % (loaded_model.metrics_names[3], score_train[3]*100))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "aG6RYmaMgvqj",
        "outputId": "41130a23-6ad2-4504-adce-003ffb2f264a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "For testing data\n",
            "accuracy: 68.38%\n",
            "recall_1: 94.12%\n",
            "auc: 77.01%\n",
            "For training data\n",
            "accuracy: 89.69%\n",
            "recall_1: 94.25%\n",
            "auc: 90.35%\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "*The performance graph after fitting the model 2nd time for better performance*\n"
      ],
      "metadata": {
        "id": "CyAIqvzRpvz4"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(score,label=\"test\")\n",
        "plt.plot(score_train,label=\"train\")\n",
        "plt.legend(ncol=2)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 286
        },
        "id": "e5vTBngIah5J",
        "outputId": "b3837692-88b0-4b7e-8d3a-9cc02e6bc0f9"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f7483b69e90>"
            ]
          },
          "metadata": {},
          "execution_count": 28
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "***Visualizing the neural network based on the density of each layer***"
      ],
      "metadata": {
        "id": "k303FiEAp5hU"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import visualkeras\n",
        "visualkeras.layered_view(loaded_model).show() # display using your system viewer\n",
        "visualkeras.layered_view(loaded_model, to_file='output.png') # write to disk\n",
        "visualkeras.layered_view(loaded_model, to_file='output.png').show() # write and show\n",
        "\n",
        "visualkeras.layered_view(loaded_model, legend=True, draw_volume=False,spacing=30)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        },
        "id": "ZOgmCPKQ1KCe",
        "outputId": "5ae655ba-284f-4782-931e-a2b7102ebcef"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<PIL.Image.Image image mode=RGBA size=355x51 at 0x7F7481642810>"
            ],
            "image/png": "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\n"
          },
          "metadata": {},
          "execution_count": 29
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from statistics import median, mode\n",
        "eval_result = loaded_model.evaluate(x_test, y_test)\n",
        "eval_result\n",
        "print(\"[test loss, test accuracy,test precision,test recall]:\", eval_result)\n",
        "predictions=loaded_model.predict(x_test)\n",
        "\n",
        "print(sum(predictions)/len(predictions))\n",
        "print(median(predictions))\n",
        "val=min(sum(predictions)/len(predictions),median(predictions))\n",
        "predictions"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3Tm6EVD7I0sE",
        "outputId": "535e9215-64d3-4fcd-8a1f-24f0a34e4250"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "15/15 [==============================] - 0s 2ms/step - loss: 0.5701 - accuracy: 0.6838 - recall_1: 0.9412 - auc: 0.7701\n",
            "[test loss, test accuracy,test precision,test recall]: [0.5701486468315125, 0.6837607026100159, 0.9411764740943909, 0.7701320648193359]\n",
            "[0.48107505]\n",
            "[0.70290196]\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0.7740873 ],\n",
              "       [0.77300334],\n",
              "       [0.7726911 ],\n",
              "       [0.06382605],\n",
              "       [0.7684273 ],\n",
              "       [0.11025763],\n",
              "       [0.42458597],\n",
              "       [0.73124427],\n",
              "       [0.4051837 ],\n",
              "       [0.7711016 ],\n",
              "       [0.18954417],\n",
              "       [0.10236096],\n",
              "       [0.12403914],\n",
              "       [0.57220364],\n",
              "       [0.77223456],\n",
              "       [0.7712364 ],\n",
              "       [0.06359369],\n",
              "       [0.68875754],\n",
              "       [0.7716042 ],\n",
              "       [0.77113914],\n",
              "       [0.771732  ],\n",
              "       [0.77005726],\n",
              "       [0.7725587 ],\n",
              "       [0.34189022],\n",
              "       [0.06694937],\n",
              "       [0.06386906],\n",
              "       [0.6002517 ],\n",
              "       [0.14342749],\n",
              "       [0.29351526],\n",
              "       [0.770697  ],\n",
              "       [0.7696937 ],\n",
              "       [0.19322237],\n",
              "       [0.77098346],\n",
              "       [0.27275506],\n",
              "       [0.06386638],\n",
              "       [0.7709289 ],\n",
              "       [0.7536638 ],\n",
              "       [0.63343394],\n",
              "       [0.06384102],\n",
              "       [0.7715918 ],\n",
              "       [0.07643488],\n",
              "       [0.7651967 ],\n",
              "       [0.18899632],\n",
              "       [0.77050716],\n",
              "       [0.7708342 ],\n",
              "       [0.06382596],\n",
              "       [0.25834078],\n",
              "       [0.06382605],\n",
              "       [0.7704178 ],\n",
              "       [0.06360275],\n",
              "       [0.7715425 ],\n",
              "       [0.08966723],\n",
              "       [0.704426  ],\n",
              "       [0.7711112 ],\n",
              "       [0.7522285 ],\n",
              "       [0.77309793],\n",
              "       [0.06408364],\n",
              "       [0.32404804],\n",
              "       [0.76347244],\n",
              "       [0.5104847 ],\n",
              "       [0.7730415 ],\n",
              "       [0.6629945 ],\n",
              "       [0.7654689 ],\n",
              "       [0.06382635],\n",
              "       [0.3590732 ],\n",
              "       [0.06393936],\n",
              "       [0.77353084],\n",
              "       [0.06382585],\n",
              "       [0.77100265],\n",
              "       [0.77256835],\n",
              "       [0.2694533 ],\n",
              "       [0.06382605],\n",
              "       [0.06396112],\n",
              "       [0.7715385 ],\n",
              "       [0.33625907],\n",
              "       [0.773083  ],\n",
              "       [0.0786078 ],\n",
              "       [0.77043676],\n",
              "       [0.07785928],\n",
              "       [0.76304424],\n",
              "       [0.7722063 ],\n",
              "       [0.77202624],\n",
              "       [0.56596726],\n",
              "       [0.23497644],\n",
              "       [0.1507211 ],\n",
              "       [0.77103424],\n",
              "       [0.644528  ],\n",
              "       [0.06382751],\n",
              "       [0.07376412],\n",
              "       [0.06383938],\n",
              "       [0.7710888 ],\n",
              "       [0.77228296],\n",
              "       [0.13635647],\n",
              "       [0.06390309],\n",
              "       [0.06382596],\n",
              "       [0.06408367],\n",
              "       [0.06544936],\n",
              "       [0.7710798 ],\n",
              "       [0.7719775 ],\n",
              "       [0.22895929],\n",
              "       [0.14061639],\n",
              "       [0.2942125 ],\n",
              "       [0.06419203],\n",
              "       [0.77284163],\n",
              "       [0.1188468 ],\n",
              "       [0.06741369],\n",
              "       [0.06382778],\n",
              "       [0.07491744],\n",
              "       [0.7713944 ],\n",
              "       [0.7709287 ],\n",
              "       [0.06447744],\n",
              "       [0.5970878 ],\n",
              "       [0.7717138 ],\n",
              "       [0.7725168 ],\n",
              "       [0.32998216],\n",
              "       [0.6678448 ],\n",
              "       [0.06410685],\n",
              "       [0.06397623],\n",
              "       [0.7711373 ],\n",
              "       [0.77000076],\n",
              "       [0.08770531],\n",
              "       [0.08939803],\n",
              "       [0.06393549],\n",
              "       [0.76838136],\n",
              "       [0.06383675],\n",
              "       [0.36227888],\n",
              "       [0.06361684],\n",
              "       [0.15037316],\n",
              "       [0.7717054 ],\n",
              "       [0.06384295],\n",
              "       [0.77117944],\n",
              "       [0.77177453],\n",
              "       [0.07026076],\n",
              "       [0.7668514 ],\n",
              "       [0.46161878],\n",
              "       [0.29980856],\n",
              "       [0.77059865],\n",
              "       [0.773825  ],\n",
              "       [0.65240806],\n",
              "       [0.77232254],\n",
              "       [0.06382757],\n",
              "       [0.6702944 ],\n",
              "       [0.7698462 ],\n",
              "       [0.76845753],\n",
              "       [0.7725347 ],\n",
              "       [0.7546431 ],\n",
              "       [0.77388847],\n",
              "       [0.07492524],\n",
              "       [0.62941766],\n",
              "       [0.77475756],\n",
              "       [0.7656775 ],\n",
              "       [0.7662866 ],\n",
              "       [0.14668885],\n",
              "       [0.71639574],\n",
              "       [0.772341  ],\n",
              "       [0.5253647 ],\n",
              "       [0.76996124],\n",
              "       [0.7718027 ],\n",
              "       [0.6390517 ],\n",
              "       [0.7709299 ],\n",
              "       [0.36751556],\n",
              "       [0.77034867],\n",
              "       [0.06458291],\n",
              "       [0.06480575],\n",
              "       [0.0638659 ],\n",
              "       [0.7726102 ],\n",
              "       [0.77136564],\n",
              "       [0.07189941],\n",
              "       [0.7709517 ],\n",
              "       [0.7342669 ],\n",
              "       [0.0638929 ],\n",
              "       [0.6814678 ],\n",
              "       [0.06386951],\n",
              "       [0.08536091],\n",
              "       [0.07346293],\n",
              "       [0.77045286],\n",
              "       [0.7703904 ],\n",
              "       [0.06388366],\n",
              "       [0.06405613],\n",
              "       [0.7723402 ],\n",
              "       [0.06385303],\n",
              "       [0.7724136 ],\n",
              "       [0.7718444 ],\n",
              "       [0.06422845],\n",
              "       [0.06459749],\n",
              "       [0.06431314],\n",
              "       [0.7728219 ],\n",
              "       [0.77202976],\n",
              "       [0.16115686],\n",
              "       [0.7726579 ],\n",
              "       [0.06409243],\n",
              "       [0.06395632],\n",
              "       [0.77198464],\n",
              "       [0.07396579],\n",
              "       [0.77164745],\n",
              "       [0.65304834],\n",
              "       [0.7700672 ],\n",
              "       [0.28881046],\n",
              "       [0.17318192],\n",
              "       [0.77251494],\n",
              "       [0.6512281 ],\n",
              "       [0.7716632 ],\n",
              "       [0.7709677 ],\n",
              "       [0.0634422 ],\n",
              "       [0.06390885],\n",
              "       [0.06391367],\n",
              "       [0.06397462],\n",
              "       [0.06386828],\n",
              "       [0.773082  ],\n",
              "       [0.7703773 ],\n",
              "       [0.06393105],\n",
              "       [0.7680721 ],\n",
              "       [0.19308683],\n",
              "       [0.0638321 ],\n",
              "       [0.06389558],\n",
              "       [0.772053  ],\n",
              "       [0.77124584],\n",
              "       [0.7722137 ],\n",
              "       [0.06383288],\n",
              "       [0.7700481 ],\n",
              "       [0.7717353 ],\n",
              "       [0.06409243],\n",
              "       [0.102678  ],\n",
              "       [0.76236486],\n",
              "       [0.7383686 ],\n",
              "       [0.77084047],\n",
              "       [0.06656861],\n",
              "       [0.06382737],\n",
              "       [0.7697942 ],\n",
              "       [0.15093437],\n",
              "       [0.760185  ],\n",
              "       [0.74454594],\n",
              "       [0.77311933],\n",
              "       [0.7718933 ],\n",
              "       [0.77020264],\n",
              "       [0.09601468],\n",
              "       [0.7724862 ],\n",
              "       [0.44481695],\n",
              "       [0.71529514],\n",
              "       [0.08812159],\n",
              "       [0.06385782],\n",
              "       [0.77192354],\n",
              "       [0.7715704 ],\n",
              "       [0.7700116 ],\n",
              "       [0.77067196],\n",
              "       [0.77127874],\n",
              "       [0.74375904],\n",
              "       [0.6781807 ],\n",
              "       [0.6130197 ],\n",
              "       [0.06382734],\n",
              "       [0.7745842 ],\n",
              "       [0.76628447],\n",
              "       [0.701378  ],\n",
              "       [0.06381851],\n",
              "       [0.06413338],\n",
              "       [0.73065245],\n",
              "       [0.45189166],\n",
              "       [0.77197874],\n",
              "       [0.06384102],\n",
              "       [0.77292764],\n",
              "       [0.7714804 ],\n",
              "       [0.77397096],\n",
              "       [0.771475  ],\n",
              "       [0.06422663],\n",
              "       [0.07392988],\n",
              "       [0.06383678],\n",
              "       [0.06374338],\n",
              "       [0.57388914],\n",
              "       [0.7719703 ],\n",
              "       [0.18183863],\n",
              "       [0.7727788 ],\n",
              "       [0.7727204 ],\n",
              "       [0.7732547 ],\n",
              "       [0.77075404],\n",
              "       [0.06342667],\n",
              "       [0.77166903],\n",
              "       [0.7726872 ],\n",
              "       [0.06528679],\n",
              "       [0.51419467],\n",
              "       [0.77309215],\n",
              "       [0.77163434],\n",
              "       [0.7720233 ],\n",
              "       [0.19755635],\n",
              "       [0.7720809 ],\n",
              "       [0.77102363],\n",
              "       [0.77108914],\n",
              "       [0.06392765],\n",
              "       [0.7702193 ],\n",
              "       [0.77168334],\n",
              "       [0.6252311 ],\n",
              "       [0.09525314],\n",
              "       [0.7701328 ],\n",
              "       [0.7715381 ],\n",
              "       [0.11696342],\n",
              "       [0.7522285 ],\n",
              "       [0.77147925],\n",
              "       [0.7727709 ],\n",
              "       [0.10580555],\n",
              "       [0.13449791],\n",
              "       [0.7704524 ],\n",
              "       [0.06393018],\n",
              "       [0.06383958],\n",
              "       [0.23242715],\n",
              "       [0.6191399 ],\n",
              "       [0.75004596],\n",
              "       [0.06525052],\n",
              "       [0.773058  ],\n",
              "       [0.06382814],\n",
              "       [0.77390134],\n",
              "       [0.0639458 ],\n",
              "       [0.45237958],\n",
              "       [0.6928273 ],\n",
              "       [0.7695819 ],\n",
              "       [0.06354135],\n",
              "       [0.76327205],\n",
              "       [0.77289623],\n",
              "       [0.7723576 ],\n",
              "       [0.5086187 ],\n",
              "       [0.06663337],\n",
              "       [0.6844096 ],\n",
              "       [0.06363854],\n",
              "       [0.06382585],\n",
              "       [0.06385124],\n",
              "       [0.0638786 ],\n",
              "       [0.06367707],\n",
              "       [0.2950405 ],\n",
              "       [0.7714155 ],\n",
              "       [0.77535176],\n",
              "       [0.26925617],\n",
              "       [0.77312636],\n",
              "       [0.07553285],\n",
              "       [0.7647646 ],\n",
              "       [0.7708275 ],\n",
              "       [0.7605673 ],\n",
              "       [0.77092   ],\n",
              "       [0.08828917],\n",
              "       [0.7696643 ],\n",
              "       [0.77126276],\n",
              "       [0.7715158 ],\n",
              "       [0.06810826],\n",
              "       [0.77055764],\n",
              "       [0.74593383],\n",
              "       [0.11194131],\n",
              "       [0.7591683 ],\n",
              "       [0.7698357 ],\n",
              "       [0.17832255],\n",
              "       [0.77082324],\n",
              "       [0.77163166],\n",
              "       [0.06646749],\n",
              "       [0.7736088 ],\n",
              "       [0.31729206],\n",
              "       [0.771871  ],\n",
              "       [0.06407684],\n",
              "       [0.7545485 ],\n",
              "       [0.7722925 ],\n",
              "       [0.06387532],\n",
              "       [0.06355858],\n",
              "       [0.7717126 ],\n",
              "       [0.771376  ],\n",
              "       [0.7697235 ],\n",
              "       [0.77206385],\n",
              "       [0.76746166],\n",
              "       [0.7714535 ],\n",
              "       [0.7727021 ],\n",
              "       [0.08301702],\n",
              "       [0.06385911],\n",
              "       [0.7722261 ],\n",
              "       [0.76974523],\n",
              "       [0.18452078],\n",
              "       [0.4786811 ],\n",
              "       [0.2569775 ],\n",
              "       [0.77201486],\n",
              "       [0.06383291],\n",
              "       [0.06382596],\n",
              "       [0.06420323],\n",
              "       [0.7541579 ],\n",
              "       [0.06944177],\n",
              "       [0.7566907 ],\n",
              "       [0.77160853],\n",
              "       [0.07721677],\n",
              "       [0.77073294],\n",
              "       [0.7716513 ],\n",
              "       [0.6380344 ],\n",
              "       [0.06382605],\n",
              "       [0.7713779 ],\n",
              "       [0.7643783 ],\n",
              "       [0.77111435],\n",
              "       [0.77277714],\n",
              "       [0.07121518],\n",
              "       [0.06385016],\n",
              "       [0.77118456],\n",
              "       [0.77209723],\n",
              "       [0.37191552],\n",
              "       [0.6869005 ],\n",
              "       [0.73829365],\n",
              "       [0.7649499 ],\n",
              "       [0.7452332 ],\n",
              "       [0.1429286 ],\n",
              "       [0.1046367 ],\n",
              "       [0.7729795 ],\n",
              "       [0.06387776],\n",
              "       [0.7730119 ],\n",
              "       [0.77398014],\n",
              "       [0.6142054 ],\n",
              "       [0.06361279],\n",
              "       [0.08986095],\n",
              "       [0.06383345],\n",
              "       [0.7714826 ],\n",
              "       [0.7708748 ],\n",
              "       [0.77208304],\n",
              "       [0.7207312 ],\n",
              "       [0.7731635 ],\n",
              "       [0.06411025],\n",
              "       [0.77149725],\n",
              "       [0.771587  ],\n",
              "       [0.77119243],\n",
              "       [0.76971155],\n",
              "       [0.32941484],\n",
              "       [0.14981088],\n",
              "       [0.7105553 ],\n",
              "       [0.06363627],\n",
              "       [0.7711928 ],\n",
              "       [0.7727368 ],\n",
              "       [0.7720854 ],\n",
              "       [0.5248836 ],\n",
              "       [0.77041376],\n",
              "       [0.7054326 ],\n",
              "       [0.06355953],\n",
              "       [0.06363252],\n",
              "       [0.06431621],\n",
              "       [0.06438076],\n",
              "       [0.06385738],\n",
              "       [0.7738129 ],\n",
              "       [0.67904276],\n",
              "       [0.56571895],\n",
              "       [0.7723586 ],\n",
              "       [0.7701516 ],\n",
              "       [0.06387439],\n",
              "       [0.7731252 ],\n",
              "       [0.7612122 ],\n",
              "       [0.7729145 ],\n",
              "       [0.77190804],\n",
              "       [0.13217491],\n",
              "       [0.77136195],\n",
              "       [0.77168953],\n",
              "       [0.7725508 ],\n",
              "       [0.06382746],\n",
              "       [0.07993171],\n",
              "       [0.7708075 ],\n",
              "       [0.7482632 ],\n",
              "       [0.0942173 ],\n",
              "       [0.06361526],\n",
              "       [0.7729449 ],\n",
              "       [0.771865  ],\n",
              "       [0.06355679],\n",
              "       [0.7709868 ],\n",
              "       [0.18209815],\n",
              "       [0.06706029],\n",
              "       [0.4222301 ],\n",
              "       [0.06406239],\n",
              "       [0.77081954],\n",
              "       [0.7705425 ],\n",
              "       [0.06395474],\n",
              "       [0.06396315],\n",
              "       [0.7649847 ],\n",
              "       [0.7705697 ],\n",
              "       [0.06425828],\n",
              "       [0.06439468]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 30
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(predictions.shape)\n",
        "seq_predictions=np.transpose(predictions)[0]  # transformation to get (n,)\n",
        "print(seq_predictions.shape)  # now the shape is (n,)\n",
        "# Applying transformation to get binary values predictions with 0.5 as thresold\n",
        "seq_predictions = list(map(lambda x: 0 if x< val else 1, seq_predictions))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZJ72SnoEJdZj",
        "outputId": "2916a09f-42a6-4593-b0a7-a82a3a0ec7b8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(468, 1)\n",
            "(468,)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "***The working of the dataset on the testing data and comparing it with its label gives us the confusion matrix which has a better recall. Recall is an important factor as higher recall implies that the model can predict plagiarised assignments almost 90% of the times***"
      ],
      "metadata": {
        "id": "pG1An0oyqEIg"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "%matplotlib inline\n",
        "from sklearn.metrics import confusion_matrix\n",
        "import itertools\n",
        "import matplotlib.pyplot as plt"
      ],
      "metadata": {
        "id": "devPsSkth2jC"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "cm = confusion_matrix(y_test,seq_predictions)"
      ],
      "metadata": {
        "id": "OGMjOwELJFEu"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def plot_confusion_matrix(cm, classes,\n",
        "                        normalize=False,\n",
        "                        title='Confusion matrix',\n",
        "                        cmap=plt.cm.Blues):\n",
        "    \"\"\"\n",
        "    This function prints and plots the confusion matrix.\n",
        "    Normalization can be applied by setting `normalize=True`.\n",
        "    \"\"\"\n",
        "    plt.imshow(cm, interpolation='nearest', cmap=cmap)\n",
        "    plt.title(title)\n",
        "    plt.colorbar()\n",
        "    tick_marks = np.arange(len(classes))\n",
        "    plt.xticks(tick_marks, classes, rotation=45)\n",
        "    plt.yticks(tick_marks, classes)\n",
        "\n",
        "    if normalize:\n",
        "        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
        "        print(\"Normalized confusion matrix\")\n",
        "    else:\n",
        "        print('Confusion matrix, without normalization')\n",
        "\n",
        "    print(cm)\n",
        "\n",
        "    thresh = cm.max() / 2.\n",
        "    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
        "        plt.text(j, i, cm[i, j],\n",
        "            horizontalalignment=\"center\",\n",
        "            color=\"white\" if cm[i, j] > thresh else \"black\")\n",
        "\n",
        "    plt.tight_layout()\n",
        "    plt.ylabel('True label')\n",
        "    plt.xlabel('Predicted label')"
      ],
      "metadata": {
        "id": "SQqMKpLbJJpq"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "cm_plot_labels = ['No Plagiarism','Plagiarism']\n",
        "plot_confusion_matrix(cm=cm, classes=cm_plot_labels, title='Confusion Matrix')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 368
        },
        "id": "O3-9ubU7JSNX",
        "outputId": "266f657e-1992-4b2a-8dbc-4433cdf0fca0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Confusion matrix, without normalization\n",
            "[[192 140]\n",
            " [  8 128]]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(\"Accuracy\",sklearn.metrics.accuracy_score(y_test,seq_predictions))\n",
        "print(\"Recall\",sklearn.metrics.recall_score(y_test,seq_predictions))\n",
        "print(\"ROC AUC\",sklearn.metrics.roc_auc_score(y_test,seq_predictions))\n",
        "print(\"F1\",sklearn.metrics.f1_score(y_test,seq_predictions))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pHZHVen6O-4P",
        "outputId": "0ea2c787-0fdc-482e-eb5a-078e1a5d17c8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy 0.6837606837606838\n",
            "Recall 0.9411764705882353\n",
            "ROC AUC 0.7597448618001418\n",
            "F1 0.6336633663366337\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "***Testing on the B2017 dataset to generate the output***"
      ],
      "metadata": {
        "id": "EcxvmWSqCvh5"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from numpy import asarray\n",
        "# values=[\n",
        "# 7216624.0,2786.0,1682.0,810.0,499.0,30029.0,18276.0,24492.0,1321.0,0.00542927,10928.0,469.0,0.616049383,3.585377013    \n",
        "#    ]\n",
        "test_2017=pd.read_csv('/content/sample_data/plagiarism_student_dataset_test_B2017.csv',index_col=0)\n",
        "test_2017.index.names=[None]\n",
        "test_2017.fillna(0,inplace=True)\n",
        "test_2017.describe()\n",
        "test_2017=scaler.fit_transform(test_2017)\n",
        "# values=np.array(values).reshape(-1,1)\n",
        "# values=scaler.fit_transform(values)\n",
        "# val=(max(values)+min(values))/2\n",
        "# print(val)\n",
        "# print(values)\n",
        "# input_array = np.array([1,2,3,4,5,6,7,8,9,10,11,12,13,14])\n",
        "# values = np.expand_dims(values,axis=0)\n",
        "label=loaded_model.predict(test_2017)\n",
        "val=min(sum(label)/len(label),median(label))\n",
        "print(label)\n",
        "for i in label:\n",
        "  print(\"Plagiarised\" if i>val else \"Non Plagiarised\")\n",
        "print(val)\n",
        "print(label.dtype)\n",
        "# sns.countplot(label)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pZ66wre8QGVa",
        "outputId": "eb84a455-15ba-4868-92ac-41d382b092e7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[0.8017603 ]\n",
            " [0.808432  ]\n",
            " [0.04523888]\n",
            " [0.7860495 ]\n",
            " [0.80311686]\n",
            " [0.80431515]\n",
            " [0.3904968 ]\n",
            " [0.80883235]\n",
            " [0.7916964 ]\n",
            " [0.04534081]\n",
            " [0.8050811 ]\n",
            " [0.04878539]\n",
            " [0.8033142 ]\n",
            " [0.04451224]\n",
            " [0.8033899 ]\n",
            " [0.04543582]\n",
            " [0.8066152 ]\n",
            " [0.80282193]\n",
            " [0.046693  ]\n",
            " [0.79984903]\n",
            " [0.8031783 ]\n",
            " [0.804325  ]\n",
            " [0.04523215]\n",
            " [0.04527128]\n",
            " [0.04793254]\n",
            " [0.04547253]\n",
            " [0.80375165]\n",
            " [0.04525322]\n",
            " [0.8055739 ]\n",
            " [0.16563755]\n",
            " [0.76106954]\n",
            " [0.80432403]\n",
            " [0.8064474 ]\n",
            " [0.06870583]\n",
            " [0.808813  ]\n",
            " [0.13719201]\n",
            " [0.80354464]\n",
            " [0.04523373]\n",
            " [0.80250764]\n",
            " [0.04523358]\n",
            " [0.8004261 ]\n",
            " [0.7796631 ]\n",
            " [0.04669365]\n",
            " [0.76801646]\n",
            " [0.05586854]\n",
            " [0.80447626]\n",
            " [0.79047966]\n",
            " [0.7854789 ]\n",
            " [0.80305576]\n",
            " [0.8080353 ]\n",
            " [0.802177  ]\n",
            " [0.8055687 ]\n",
            " [0.80223584]\n",
            " [0.80342305]\n",
            " [0.75898975]\n",
            " [0.7137796 ]\n",
            " [0.8035396 ]\n",
            " [0.540498  ]\n",
            " [0.0452337 ]\n",
            " [0.04526219]\n",
            " [0.8049438 ]\n",
            " [0.80621463]\n",
            " [0.8037942 ]\n",
            " [0.7224908 ]\n",
            " [0.04524791]\n",
            " [0.04538193]\n",
            " [0.04722577]\n",
            " [0.13494524]\n",
            " [0.4730414 ]\n",
            " [0.8066945 ]\n",
            " [0.8038046 ]\n",
            " [0.44422728]\n",
            " [0.04523316]\n",
            " [0.36013323]\n",
            " [0.80972147]\n",
            " [0.80697644]\n",
            " [0.6362988 ]\n",
            " [0.04701835]\n",
            " [0.16995463]\n",
            " [0.06385615]\n",
            " [0.0446699 ]\n",
            " [0.7978262 ]\n",
            " [0.8007282 ]\n",
            " [0.04517871]\n",
            " [0.8049431 ]\n",
            " [0.80655575]\n",
            " [0.8085275 ]\n",
            " [0.8029459 ]\n",
            " [0.13114989]\n",
            " [0.80554974]\n",
            " [0.8039808 ]\n",
            " [0.8031652 ]\n",
            " [0.8055092 ]\n",
            " [0.04632732]\n",
            " [0.8013941 ]\n",
            " [0.04532427]\n",
            " [0.17831424]\n",
            " [0.8027172 ]\n",
            " [0.80282485]\n",
            " [0.04532123]\n",
            " [0.04497436]\n",
            " [0.52383745]\n",
            " [0.80472285]\n",
            " [0.06635639]\n",
            " [0.04524195]\n",
            " [0.36693245]\n",
            " [0.05834717]\n",
            " [0.08894435]\n",
            " [0.7967036 ]\n",
            " [0.04525203]\n",
            " [0.80452085]\n",
            " [0.04523543]\n",
            " [0.80765283]\n",
            " [0.04535094]\n",
            " [0.0452356 ]\n",
            " [0.8023087 ]\n",
            " [0.04612267]\n",
            " [0.41582412]\n",
            " [0.80841196]\n",
            " [0.04523718]\n",
            " [0.0457029 ]\n",
            " [0.7471615 ]\n",
            " [0.7518604 ]\n",
            " [0.8035153 ]\n",
            " [0.5725651 ]\n",
            " [0.8054254 ]\n",
            " [0.80164444]\n",
            " [0.04523373]\n",
            " [0.04453668]\n",
            " [0.8052919 ]\n",
            " [0.05455714]\n",
            " [0.8060198 ]\n",
            " [0.045569  ]\n",
            " [0.04534906]\n",
            " [0.8029714 ]\n",
            " [0.80285525]\n",
            " [0.04525065]\n",
            " [0.8056897 ]\n",
            " [0.8069925 ]\n",
            " [0.05971196]\n",
            " [0.802747  ]\n",
            " [0.8030307 ]\n",
            " [0.06532332]\n",
            " [0.04785728]\n",
            " [0.8026586 ]\n",
            " [0.04523373]\n",
            " [0.8066062 ]\n",
            " [0.04783526]\n",
            " [0.04497153]\n",
            " [0.80386454]\n",
            " [0.04524702]\n",
            " [0.04533005]\n",
            " [0.8017185 ]\n",
            " [0.04541212]\n",
            " [0.07353586]\n",
            " [0.8040838 ]\n",
            " [0.80565   ]\n",
            " [0.04591432]\n",
            " [0.81106687]\n",
            " [0.80126876]\n",
            " [0.05065268]\n",
            " [0.6508888 ]\n",
            " [0.12225255]\n",
            " [0.04523361]\n",
            " [0.04523373]\n",
            " [0.6739511 ]\n",
            " [0.8041506 ]\n",
            " [0.05041286]\n",
            " [0.8049221 ]\n",
            " [0.59689564]\n",
            " [0.04538733]\n",
            " [0.0452337 ]\n",
            " [0.6699089 ]\n",
            " [0.04523373]\n",
            " [0.04534385]\n",
            " [0.7774653 ]\n",
            " [0.8025693 ]\n",
            " [0.06465772]\n",
            " [0.04466522]\n",
            " [0.04547438]\n",
            " [0.80823475]\n",
            " [0.8035633 ]\n",
            " [0.38663262]\n",
            " [0.8124994 ]\n",
            " [0.72781384]\n",
            " [0.8014749 ]\n",
            " [0.80527174]\n",
            " [0.04586324]\n",
            " [0.80153656]\n",
            " [0.04529995]\n",
            " [0.04728091]\n",
            " [0.80384254]\n",
            " [0.04523316]\n",
            " [0.60745704]\n",
            " [0.779983  ]\n",
            " [0.8037998 ]\n",
            " [0.8039181 ]\n",
            " [0.04523361]\n",
            " [0.42732203]\n",
            " [0.04527682]\n",
            " [0.4311317 ]\n",
            " [0.6387502 ]\n",
            " [0.8054804 ]\n",
            " [0.80276966]\n",
            " [0.8081712 ]\n",
            " [0.7948977 ]\n",
            " [0.80708003]\n",
            " [0.04488221]\n",
            " [0.0452489 ]\n",
            " [0.80516946]\n",
            " [0.04638341]\n",
            " [0.8058712 ]\n",
            " [0.04697862]\n",
            " [0.8048476 ]\n",
            " [0.04525203]\n",
            " [0.7848843 ]\n",
            " [0.05832133]\n",
            " [0.29820585]\n",
            " [0.0452354 ]\n",
            " [0.7447837 ]\n",
            " [0.04524425]\n",
            " [0.7266605 ]\n",
            " [0.80421287]\n",
            " [0.80220294]\n",
            " [0.08758217]\n",
            " [0.7907303 ]\n",
            " [0.04526559]\n",
            " [0.80815476]\n",
            " [0.04524776]\n",
            " [0.7025988 ]\n",
            " [0.8052513 ]\n",
            " [0.04523373]\n",
            " [0.05459195]\n",
            " [0.04477856]\n",
            " [0.04453668]\n",
            " [0.8077613 ]\n",
            " [0.2641458 ]\n",
            " [0.1221185 ]\n",
            " [0.7563182 ]\n",
            " [0.65895057]\n",
            " [0.8096956 ]\n",
            " [0.04585084]\n",
            " [0.6694026 ]\n",
            " [0.8006488 ]\n",
            " [0.29970104]\n",
            " [0.0452337 ]\n",
            " [0.22040582]\n",
            " [0.71690464]\n",
            " [0.04981858]\n",
            " [0.04558843]\n",
            " [0.8016784 ]\n",
            " [0.8057632 ]\n",
            " [0.04525545]\n",
            " [0.05621955]\n",
            " [0.04525039]\n",
            " [0.18146548]\n",
            " [0.80425584]\n",
            " [0.80785334]\n",
            " [0.04498801]\n",
            " [0.04568896]\n",
            " [0.04523364]\n",
            " [0.80414736]\n",
            " [0.5931659 ]\n",
            " [0.04721558]\n",
            " [0.8041197 ]\n",
            " [0.8050014 ]\n",
            " [0.8033097 ]\n",
            " [0.8060302 ]\n",
            " [0.04523215]\n",
            " [0.803864  ]\n",
            " [0.8048527 ]\n",
            " [0.06551462]\n",
            " [0.04536045]\n",
            " [0.08816525]\n",
            " [0.0452337 ]\n",
            " [0.4128942 ]\n",
            " [0.80415964]\n",
            " [0.8062202 ]\n",
            " [0.7077507 ]\n",
            " [0.8049861 ]\n",
            " [0.8068533 ]\n",
            " [0.8062632 ]\n",
            " [0.3453827 ]\n",
            " [0.0452337 ]\n",
            " [0.8073888 ]\n",
            " [0.8043517 ]\n",
            " [0.8075384 ]\n",
            " [0.76910114]\n",
            " [0.04541913]\n",
            " [0.04527745]\n",
            " [0.04565045]\n",
            " [0.81160825]\n",
            " [0.37595722]\n",
            " [0.04523373]\n",
            " [0.06252244]\n",
            " [0.22932202]\n",
            " [0.04540634]\n",
            " [0.04796851]\n",
            " [0.80890155]\n",
            " [0.80082345]\n",
            " [0.04523543]\n",
            " [0.04527146]\n",
            " [0.04536059]\n",
            " [0.0470033 ]\n",
            " [0.16922739]\n",
            " [0.2156975 ]\n",
            " [0.04524872]\n",
            " [0.06016156]\n",
            " [0.7706672 ]\n",
            " [0.04528388]\n",
            " [0.8035375 ]\n",
            " [0.04526052]\n",
            " [0.43494797]\n",
            " [0.80511427]\n",
            " [0.0452337 ]\n",
            " [0.8023652 ]\n",
            " [0.8098892 ]\n",
            " [0.8031759 ]\n",
            " [0.04526389]\n",
            " [0.804428  ]\n",
            " [0.80257094]\n",
            " [0.81356895]\n",
            " [0.04529697]\n",
            " [0.8035098 ]\n",
            " [0.802824  ]\n",
            " [0.8058679 ]\n",
            " [0.04523361]\n",
            " [0.04524028]\n",
            " [0.80523926]\n",
            " [0.3291266 ]\n",
            " [0.04524365]\n",
            " [0.25313973]\n",
            " [0.06819427]\n",
            " [0.46645704]\n",
            " [0.6700197 ]\n",
            " [0.04530427]\n",
            " [0.8062569 ]\n",
            " [0.16951445]\n",
            " [0.08113331]\n",
            " [0.04540694]\n",
            " [0.16437626]\n",
            " [0.5861905 ]\n",
            " [0.80728483]\n",
            " [0.44894776]\n",
            " [0.04524067]\n",
            " [0.70528495]\n",
            " [0.0456399 ]\n",
            " [0.04523864]\n",
            " [0.81035364]\n",
            " [0.80265886]\n",
            " [0.80167913]\n",
            " [0.04523543]\n",
            " [0.80584276]\n",
            " [0.80432796]\n",
            " [0.80484354]\n",
            " [0.17231101]\n",
            " [0.09825426]\n",
            " [0.804266  ]\n",
            " [0.6363001 ]\n",
            " [0.80222905]\n",
            " [0.0483807 ]\n",
            " [0.04525143]\n",
            " [0.8032677 ]\n",
            " [0.04523724]\n",
            " [0.045237  ]\n",
            " [0.77805877]\n",
            " [0.8027918 ]\n",
            " [0.07718021]\n",
            " [0.7936913 ]\n",
            " [0.73108584]\n",
            " [0.09911031]\n",
            " [0.8038602 ]\n",
            " [0.04562104]\n",
            " [0.04568431]\n",
            " [0.74668324]\n",
            " [0.7554301 ]\n",
            " [0.80465114]\n",
            " [0.8110114 ]\n",
            " [0.8034897 ]\n",
            " [0.8047681 ]\n",
            " [0.57337284]\n",
            " [0.8039208 ]\n",
            " [0.04503313]\n",
            " [0.04608586]\n",
            " [0.8030854 ]\n",
            " [0.8030362 ]\n",
            " [0.5302206 ]\n",
            " [0.80439526]\n",
            " [0.80752313]\n",
            " [0.7989644 ]\n",
            " [0.8034687 ]\n",
            " [0.74212825]\n",
            " [0.0452337 ]\n",
            " [0.8059827 ]\n",
            " [0.04812202]\n",
            " [0.8062347 ]\n",
            " [0.61400366]\n",
            " [0.8056174 ]\n",
            " [0.06543317]\n",
            " [0.80585015]\n",
            " [0.04547083]\n",
            " [0.80423504]\n",
            " [0.8053962 ]\n",
            " [0.04531479]\n",
            " [0.8034551 ]\n",
            " [0.04548118]\n",
            " [0.05640119]\n",
            " [0.80347985]\n",
            " [0.8030994 ]\n",
            " [0.04562831]\n",
            " [0.8048292 ]\n",
            " [0.45202088]\n",
            " [0.806142  ]\n",
            " [0.8054162 ]\n",
            " [0.49400756]\n",
            " [0.8059517 ]\n",
            " [0.78541815]\n",
            " [0.5339458 ]\n",
            " [0.80367005]\n",
            " [0.04527643]\n",
            " [0.8029381 ]\n",
            " [0.04524702]\n",
            " [0.7422487 ]\n",
            " [0.80812657]\n",
            " [0.04585242]\n",
            " [0.4272553 ]\n",
            " [0.47913364]\n",
            " [0.80360913]\n",
            " [0.8059502 ]\n",
            " [0.81034315]\n",
            " [0.8063139 ]\n",
            " [0.8041357 ]\n",
            " [0.6974999 ]\n",
            " [0.80530465]\n",
            " [0.8048359 ]\n",
            " [0.04523864]\n",
            " [0.07923827]\n",
            " [0.8048081 ]\n",
            " [0.04545277]\n",
            " [0.0458481 ]\n",
            " [0.05002624]\n",
            " [0.04664481]\n",
            " [0.8015671 ]\n",
            " [0.8044083 ]\n",
            " [0.7979542 ]\n",
            " [0.16889673]\n",
            " [0.8015822 ]\n",
            " [0.04523867]\n",
            " [0.7702071 ]\n",
            " [0.04679868]\n",
            " [0.7569699 ]\n",
            " [0.04569018]\n",
            " [0.80802613]\n",
            " [0.04524195]\n",
            " [0.805681  ]\n",
            " [0.80524653]\n",
            " [0.80579853]\n",
            " [0.8023795 ]\n",
            " [0.04491866]\n",
            " [0.0452354 ]\n",
            " [0.8036629 ]\n",
            " [0.04525715]\n",
            " [0.74828583]\n",
            " [0.80735564]\n",
            " [0.7853675 ]\n",
            " [0.80291337]\n",
            " [0.56840634]\n",
            " [0.04607633]\n",
            " [0.81484187]\n",
            " [0.04524365]\n",
            " [0.8027314 ]\n",
            " [0.04585412]\n",
            " [0.04528606]\n",
            " [0.04526043]\n",
            " [0.8033049 ]\n",
            " [0.04517943]\n",
            " [0.80260265]\n",
            " [0.8058573 ]\n",
            " [0.04527509]\n",
            " [0.48895487]\n",
            " [0.8010471 ]\n",
            " [0.10971057]\n",
            " [0.80390817]\n",
            " [0.78936577]\n",
            " [0.8018599 ]\n",
            " [0.8057542 ]\n",
            " [0.6687467 ]\n",
            " [0.80460054]\n",
            " [0.805358  ]\n",
            " [0.8074147 ]\n",
            " [0.15024775]\n",
            " [0.04531682]\n",
            " [0.07934886]\n",
            " [0.8040279 ]\n",
            " [0.04447028]\n",
            " [0.8052317 ]\n",
            " [0.04612091]\n",
            " [0.04726321]\n",
            " [0.4446568 ]\n",
            " [0.04527351]\n",
            " [0.8022511 ]\n",
            " [0.8060528 ]\n",
            " [0.1299322 ]\n",
            " [0.78315705]\n",
            " [0.8072813 ]\n",
            " [0.04528546]\n",
            " [0.80543697]\n",
            " [0.80673885]\n",
            " [0.0452331 ]\n",
            " [0.05613142]\n",
            " [0.80307204]\n",
            " [0.11628002]\n",
            " [0.7777393 ]\n",
            " [0.0452337 ]\n",
            " [0.05072275]\n",
            " [0.1531074 ]\n",
            " [0.8048691 ]\n",
            " [0.04544982]\n",
            " [0.8054954 ]\n",
            " [0.04523361]\n",
            " [0.6952056 ]\n",
            " [0.79762256]\n",
            " [0.04562983]\n",
            " [0.80157745]\n",
            " [0.80955756]\n",
            " [0.8035694 ]\n",
            " [0.04467893]\n",
            " [0.0452337 ]\n",
            " [0.80204177]\n",
            " [0.80190444]\n",
            " [0.80698776]\n",
            " [0.8064159 ]\n",
            " [0.1410892 ]\n",
            " [0.0467779 ]\n",
            " [0.8067104 ]\n",
            " [0.35133433]\n",
            " [0.08728427]\n",
            " [0.4227986 ]\n",
            " [0.8039565 ]\n",
            " [0.04636669]\n",
            " [0.7594925 ]\n",
            " [0.74379975]\n",
            " [0.8039304 ]\n",
            " [0.8021497 ]\n",
            " [0.04527745]\n",
            " [0.04535514]\n",
            " [0.04524028]\n",
            " [0.7476809 ]\n",
            " [0.04523361]\n",
            " [0.1670838 ]\n",
            " [0.04525506]\n",
            " [0.8025464 ]\n",
            " [0.706859  ]\n",
            " [0.7832002 ]\n",
            " [0.75811386]\n",
            " [0.04672602]\n",
            " [0.80520993]\n",
            " [0.8044821 ]\n",
            " [0.80464864]\n",
            " [0.2842099 ]\n",
            " [0.80373687]\n",
            " [0.6699283 ]\n",
            " [0.8027825 ]\n",
            " [0.8045443 ]\n",
            " [0.21462512]\n",
            " [0.80565774]\n",
            " [0.04524702]\n",
            " [0.04523373]\n",
            " [0.04523373]\n",
            " [0.8055179 ]\n",
            " [0.8078059 ]\n",
            " [0.8057524 ]\n",
            " [0.8083385 ]\n",
            " [0.746125  ]\n",
            " [0.8038069 ]\n",
            " [0.04572973]\n",
            " [0.15352142]\n",
            " [0.80831045]\n",
            " [0.6556597 ]\n",
            " [0.80682063]\n",
            " [0.7931764 ]\n",
            " [0.08473542]\n",
            " [0.04537964]\n",
            " [0.06402168]\n",
            " [0.80436957]\n",
            " [0.669383  ]\n",
            " [0.80966055]\n",
            " [0.80053836]\n",
            " [0.8023118 ]\n",
            " [0.04523581]\n",
            " [0.04524365]\n",
            " [0.401573  ]\n",
            " [0.803661  ]\n",
            " [0.04524601]\n",
            " [0.04587889]\n",
            " [0.81181276]\n",
            " [0.80416024]\n",
            " [0.80640614]\n",
            " [0.04524535]\n",
            " [0.8026272 ]\n",
            " [0.0452424 ]\n",
            " [0.8039565 ]\n",
            " [0.08116987]\n",
            " [0.8065711 ]\n",
            " [0.8056884 ]\n",
            " [0.80376756]\n",
            " [0.04746032]\n",
            " [0.05372971]\n",
            " [0.803027  ]\n",
            " [0.8013737 ]\n",
            " [0.04527882]\n",
            " [0.60981965]\n",
            " [0.04525039]\n",
            " [0.80599415]\n",
            " [0.8074684 ]\n",
            " [0.8046292 ]\n",
            " [0.46261525]\n",
            " [0.81039375]\n",
            " [0.80387855]\n",
            " [0.80303895]\n",
            " [0.80317557]\n",
            " [0.80866385]\n",
            " [0.8033196 ]\n",
            " [0.8057535 ]\n",
            " [0.8016969 ]\n",
            " [0.7896069 ]\n",
            " [0.80313027]\n",
            " [0.80570537]\n",
            " [0.10560882]\n",
            " [0.04524872]\n",
            " [0.8003094 ]\n",
            " [0.6713926 ]\n",
            " [0.8036837 ]\n",
            " [0.04543498]\n",
            " [0.8061929 ]\n",
            " [0.6484726 ]\n",
            " [0.04590389]\n",
            " [0.81243044]\n",
            " [0.04582453]\n",
            " [0.7797191 ]\n",
            " [0.04523215]\n",
            " [0.04524028]\n",
            " [0.04513252]\n",
            " [0.04523373]\n",
            " [0.04523864]\n",
            " [0.1114085 ]\n",
            " [0.8033395 ]\n",
            " [0.80702007]\n",
            " [0.7989301 ]\n",
            " [0.80509686]\n",
            " [0.04523215]\n",
            " [0.8020309 ]\n",
            " [0.80496013]\n",
            " [0.8049015 ]\n",
            " [0.7957032 ]\n",
            " [0.04563677]\n",
            " [0.8044305 ]\n",
            " [0.8014107 ]\n",
            " [0.7110501 ]\n",
            " [0.46913698]\n",
            " [0.4163994 ]\n",
            " [0.8061359 ]\n",
            " [0.04524365]\n",
            " [0.806731  ]\n",
            " [0.79848605]\n",
            " [0.6402231 ]\n",
            " [0.04523581]\n",
            " [0.80095065]\n",
            " [0.8057642 ]\n",
            " [0.79854256]\n",
            " [0.05291623]\n",
            " [0.3598364 ]\n",
            " [0.8061358 ]\n",
            " [0.8013605 ]\n",
            " [0.7933204 ]\n",
            " [0.04548648]\n",
            " [0.8049276 ]\n",
            " [0.8018515 ]\n",
            " [0.8030741 ]\n",
            " [0.80439556]\n",
            " [0.80254257]\n",
            " [0.8031726 ]\n",
            " [0.04563874]\n",
            " [0.79772365]\n",
            " [0.0452337 ]\n",
            " [0.803015  ]\n",
            " [0.80812126]\n",
            " [0.7873816 ]\n",
            " [0.04635522]\n",
            " [0.6135952 ]\n",
            " [0.7177962 ]\n",
            " [0.8054961 ]]\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Non Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "Plagiarised\n",
            "[0.48559937]\n",
            "float32\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# ***In conclusion we can say that the model after sampling performs better than the model before sampling with better recall and accuracy***"
      ],
      "metadata": {
        "id": "hK0tWLUNCylI"
      }
    }
  ],
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 2
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython2",
      "version": "2.7.6"
    },
    "colab": {
      "name": "Plagiarism_detection_in_computer_programming_assignments__after_sampling.ipynb",
      "provenance": [],
      "collapsed_sections": []
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}