Project1_Android.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Project1_Android.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "v64M0OIVxpLm"
      },
      "source": [
        "# Project1 : Android malware detection"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VsocnqRVHW_5"
      },
      "source": [
        "# Import Libraries \n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_a6gGwKJHThd"
      },
      "source": [
        "#read data\n",
        "import pandas as pd     # manipulate data, data processing, load csv file I/O (e.g. pd.read_csv)\n",
        "import numpy as np      # Linear Algebra (calculate the mean and standard deviation)\n",
        "\n",
        "#To visualize\n",
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline\n",
        "\n",
        "# style\n",
        "plt.style.use(\"fivethirtyeight\")       # Set Graphs Background style using matplotlib\n",
        "sns.set_style(\"darkgrid\")              # Set Graphs Background style using seaborn\n",
        "\n",
        "import warnings\n",
        "warnings.filterwarnings('ignore')\n",
        "\n",
        "from wordcloud import STOPWORDS"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "33AVhXdvHbtC"
      },
      "source": [
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "from sklearn.metrics import classification_report, confusion_matrix,accuracy_score,cohen_kappa_score\n",
        "\n",
        "from sklearn import tree\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "from sklearn.tree import DecisionTreeClassifier\n",
        "from sklearn.naive_bayes import MultinomialNB\n",
        "from sklearn.neighbors import KNeighborsClassifier\n",
        "from sklearn.ensemble import AdaBoostClassifier\n",
        "from lightgbm import LGBMClassifier, plot_importance\n",
        "from sklearn.naive_bayes import GaussianNB\n",
        "\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.pipeline import make_pipeline\n",
        "from sklearn.ensemble import BaggingClassifier\n",
        "\n",
        "from sklearn import datasets\n",
        "from sklearn import metrics"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nwX0JDAKHcE3"
      },
      "source": [
        "# Import Data \n",
        "\n",
        "---\n",
        "\n",
        "\n",
        "\n",
        "---\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dw53Ox7Qukbf",
        "outputId": "de63597e-7e07-4663-f854-d193072cca43"
      },
      "source": [
        "# Load the Drive helper and mount\n",
        "from google.colab import drive\n",
        "\n",
        "# This will prompt for authorization.\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "x-xCn9-1HkFu"
      },
      "source": [
        "\n",
        "ap = pd.read_csv('/content/drive/MyDrive/Colab Notebooks/ML Project/Android_Permission_Data/Android_Permission.csv')"
      ],
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "niYtuGioaDTy"
      },
      "source": [
        "# Preprocessing "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "tWlqjXRiP36r",
        "outputId": "2066cb98-50dd-4f88-8149-6efd478e04a6"
      },
      "source": [
        "# Listing Number of missing values by feature column wise\n",
        "total = ap.isnull().sum().sort_values(ascending=False)\n",
        "total = total[total > 0]\n",
        "total"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Related apps                   755\n",
              "Dangerous permissions count    204\n",
              "Description                      3\n",
              "App                              1\n",
              "dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zORdh9RLP9eV",
        "outputId": "2d8411b1-ffaa-4678-b0c0-d7f7a7a58ff5"
      },
      "source": [
        "ap.isnull().sum()"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "App                                                                      1\n",
              "Package                                                                  0\n",
              "Category                                                                 0\n",
              "Description                                                              3\n",
              "Rating                                                                   0\n",
              "                                                                        ..\n",
              "Your personal information : set alarm in alarm clock (S)                 0\n",
              "Your personal information : write Browser's history and bookmarks (D)    0\n",
              "Your personal information : write contact data (D)                       0\n",
              "Your personal information : write to user defined dictionary (S)         0\n",
              "Class                                                                    0\n",
              "Length: 184, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "w2Ubl7wYxkM8",
        "outputId": "b1d49202-1ac0-4ec2-ac0d-d038ba871ff2"
      },
      "source": [
        "ap.dropna(inplace=True)\n",
        "ap"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>App</th>\n",
              "      <th>Package</th>\n",
              "      <th>Category</th>\n",
              "      <th>Description</th>\n",
              "      <th>Rating</th>\n",
              "      <th>Number of ratings</th>\n",
              "      <th>Price</th>\n",
              "      <th>Related apps</th>\n",
              "      <th>Dangerous permissions count</th>\n",
              "      <th>Safe permissions count</th>\n",
              "      <th>Default : Access DRM content. (S)</th>\n",
              "      <th>Default : Access Email provider data (S)</th>\n",
              "      <th>Default : Access all system downloads (S)</th>\n",
              "      <th>Default : Access download manager. (S)</th>\n",
              "      <th>Default : Advanced download manager functions. (S)</th>\n",
              "      <th>Default : Audio File Access (S)</th>\n",
              "      <th>Default : Install DRM content. (S)</th>\n",
              "      <th>Default : Modify Google service configuration (S)</th>\n",
              "      <th>Default : Modify Google settings (S)</th>\n",
              "      <th>Default : Move application resources (S)</th>\n",
              "      <th>Default : Read Google settings (S)</th>\n",
              "      <th>Default : Send download notifications. (S)</th>\n",
              "      <th>Default : Voice Search Shortcuts (S)</th>\n",
              "      <th>Default : access SurfaceFlinger (S)</th>\n",
              "      <th>Default : access checkin properties (S)</th>\n",
              "      <th>Default : access the cache filesystem (S)</th>\n",
              "      <th>Default : access to passwords for Google accounts (S)</th>\n",
              "      <th>Default : act as an account authenticator (S)</th>\n",
              "      <th>Default : bind to a wallpaper (S)</th>\n",
              "      <th>Default : bind to an input method (S)</th>\n",
              "      <th>Default : change screen orientation (S)</th>\n",
              "      <th>Default : coarse (network-based) location (S)</th>\n",
              "      <th>Default : control location update notifications (S)</th>\n",
              "      <th>Default : control system backup and restore (S)</th>\n",
              "      <th>Default : delete applications (S)</th>\n",
              "      <th>Default : delete other applications' caches (S)</th>\n",
              "      <th>Default : delete other applications' data (S)</th>\n",
              "      <th>Default : directly call any phone numbers (S)</th>\n",
              "      <th>Default : directly install applications (S)</th>\n",
              "      <th>Default : disable or modify status bar (S)</th>\n",
              "      <th>...</th>\n",
              "      <th>Your accounts : access other Google services (D)</th>\n",
              "      <th>Your accounts : act as an account authenticator (D)</th>\n",
              "      <th>Your accounts : act as the AccountManagerService (S)</th>\n",
              "      <th>Your accounts : contacts data in Google accounts (D)</th>\n",
              "      <th>Your accounts : discover known accounts (S)</th>\n",
              "      <th>Your accounts : manage the accounts list (D)</th>\n",
              "      <th>Your accounts : read Google service configuration (S)</th>\n",
              "      <th>Your accounts : use the authentication credentials of an account (D)</th>\n",
              "      <th>Your accounts : view configured accounts (S)</th>\n",
              "      <th>Your location : access extra location provider commands (S)</th>\n",
              "      <th>Your location : coarse (network-based) location (D)</th>\n",
              "      <th>Your location : fine (GPS) location (D)</th>\n",
              "      <th>Your location : mock location sources for testing (D)</th>\n",
              "      <th>Your messages : Read Email attachments (D)</th>\n",
              "      <th>Your messages : Send Gmail (S)</th>\n",
              "      <th>Your messages : edit SMS or MMS (D)</th>\n",
              "      <th>Your messages : modify Gmail (D)</th>\n",
              "      <th>Your messages : read Gmail (D)</th>\n",
              "      <th>Your messages : read Gmail attachment previews (D)</th>\n",
              "      <th>Your messages : read SMS or MMS (D)</th>\n",
              "      <th>Your messages : read instant messages (D)</th>\n",
              "      <th>Your messages : receive MMS (D)</th>\n",
              "      <th>Your messages : receive SMS (D)</th>\n",
              "      <th>Your messages : receive WAP (D)</th>\n",
              "      <th>Your messages : send SMS-received broadcast (S)</th>\n",
              "      <th>Your messages : send WAP-PUSH-received broadcast (S)</th>\n",
              "      <th>Your messages : write instant messages (D)</th>\n",
              "      <th>Your personal information : add or modify calendar events and send email to guests (D)</th>\n",
              "      <th>Your personal information : choose widgets (S)</th>\n",
              "      <th>Your personal information : read Browser's history and bookmarks (D)</th>\n",
              "      <th>Your personal information : read calendar events (D)</th>\n",
              "      <th>Your personal information : read contact data (D)</th>\n",
              "      <th>Your personal information : read sensitive log data (D)</th>\n",
              "      <th>Your personal information : read user defined dictionary (D)</th>\n",
              "      <th>Your personal information : retrieve system internal state (S)</th>\n",
              "      <th>Your personal information : set alarm in alarm clock (S)</th>\n",
              "      <th>Your personal information : write Browser's history and bookmarks (D)</th>\n",
              "      <th>Your personal information : write contact data (D)</th>\n",
              "      <th>Your personal information : write to user defined dictionary (S)</th>\n",
              "      <th>Class</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Canada Post Corporation</td>\n",
              "      <td>com.canadapost.android</td>\n",
              "      <td>Business</td>\n",
              "      <td>Canada Post Mobile App gives you access to som...</td>\n",
              "      <td>3.1</td>\n",
              "      <td>77</td>\n",
              "      <td>0.00</td>\n",
              "      <td>{com.adaffix.pub.ca.android, com.kevinquan.gas...</td>\n",
              "      <td>7.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Word Farm</td>\n",
              "      <td>com.realcasualgames.words</td>\n",
              "      <td>Brain &amp; Puzzle</td>\n",
              "      <td>Speed and strategy combine in this exciting wo...</td>\n",
              "      <td>4.3</td>\n",
              "      <td>199</td>\n",
              "      <td>0.00</td>\n",
              "      <td>{air.com.zubawing.FastWordLite, com.joybits.do...</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Fortunes of War FREE</td>\n",
              "      <td>fortunesofwar.free</td>\n",
              "      <td>Cards &amp; Casino</td>\n",
              "      <td>Fortunes of War is a fast-paced, easy to learn...</td>\n",
              "      <td>4.1</td>\n",
              "      <td>243</td>\n",
              "      <td>0.00</td>\n",
              "      <td>{com.kevinquan.condado, hu.monsta.pazaak, net....</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Better Keyboard: Avatar Purple</td>\n",
              "      <td>com.cc.betterkeyboard.skins.avatarpurple</td>\n",
              "      <td>Libraries &amp; Demo</td>\n",
              "      <td>Skin for Better Keyboard featuring a glossy fe...</td>\n",
              "      <td>3.6</td>\n",
              "      <td>2589</td>\n",
              "      <td>0.00</td>\n",
              "      <td>{eu.gdumoulin.betterandroid.skins.transparent....</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>Ms Claus Live Wallpaper</td>\n",
              "      <td>tmc.christmaslady.livewallpaper</td>\n",
              "      <td>Personalization</td>\n",
              "      <td>Ms Claus Live Wallpaper&lt;p&gt;Find more Free apps ...</td>\n",
              "      <td>4.5</td>\n",
              "      <td>16</td>\n",
              "      <td>0.00</td>\n",
              "      <td>{tmc.christmassanta.livewallpaper, tmc.winterh...</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>29994</th>\n",
              "      <td>Beechwood Hotel</td>\n",
              "      <td>smartstay.beechwood</td>\n",
              "      <td>Travel &amp; Local</td>\n",
              "      <td>Beechwood is an Android hotel app for guests, ...</td>\n",
              "      <td>5.0</td>\n",
              "      <td>2</td>\n",
              "      <td>0.00</td>\n",
              "      <td>{com.bluewaterpages, com.travelconnection.luxu...</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>29995</th>\n",
              "      <td>ZDefense: Tower Defense</td>\n",
              "      <td>mobi.zstudio.avi.full</td>\n",
              "      <td>Arcade &amp; Action</td>\n",
              "      <td>Explosive Tower Defense!&lt;p&gt;Top Tower Defense G...</td>\n",
              "      <td>4.6</td>\n",
              "      <td>56</td>\n",
              "      <td>2.99</td>\n",
              "      <td>{com.entwicklerx.swampdefense, com.digimentgam...</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>29996</th>\n",
              "      <td>Super Task Killer FREE</td>\n",
              "      <td>com.netqin.aotkiller</td>\n",
              "      <td>Tools</td>\n",
              "      <td>Get it for FREE. Quickly Kill running apps. Fr...</td>\n",
              "      <td>4.2</td>\n",
              "      <td>6101</td>\n",
              "      <td>0.00</td>\n",
              "      <td>{com.guardam.TaskKiller, com.netqin.authtoken_...</td>\n",
              "      <td>4.0</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>29997</th>\n",
              "      <td>Ambassador Hotel Taipei</td>\n",
              "      <td>VMC.Ambassadorhotel</td>\n",
              "      <td>Travel &amp; Local</td>\n",
              "      <td>Key features of this App: &lt;br&gt;1. View photos o...</td>\n",
              "      <td>3.4</td>\n",
              "      <td>5</td>\n",
              "      <td>0.00</td>\n",
              "      <td>{com.hyxen.app.NewTaipeiTravel, j2ab.android.t...</td>\n",
              "      <td>3.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>29998</th>\n",
              "      <td>Brain Ball Game</td>\n",
              "      <td>myapp.lecz.android.tiltmazes</td>\n",
              "      <td>Brain &amp; Puzzle</td>\n",
              "      <td>A ball sits in a flat tray containing one or m...</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2</td>\n",
              "      <td>0.00</td>\n",
              "      <td>{com.twistedmirror.alberiplus, com.freesoul.Mo...</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>29240 rows × 184 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "                                  App  ... Class\n",
              "0             Canada Post Corporation  ...     0\n",
              "1                           Word Farm  ...     0\n",
              "2                Fortunes of War FREE  ...     0\n",
              "3      Better Keyboard: Avatar Purple  ...     0\n",
              "5             Ms Claus Live Wallpaper  ...     0\n",
              "...                               ...  ...   ...\n",
              "29994                 Beechwood Hotel  ...     1\n",
              "29995         ZDefense: Tower Defense  ...     1\n",
              "29996          Super Task Killer FREE  ...     1\n",
              "29997         Ambassador Hotel Taipei  ...     1\n",
              "29998                 Brain Ball Game  ...     1\n",
              "\n",
              "[29240 rows x 184 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Bl4_4xLRxpZ9",
        "outputId": "7d20f803-6b18-45f5-d793-8264de402c73"
      },
      "source": [
        "ap.isnull().values.any()"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "False"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "B5Ae30OTqZGf"
      },
      "source": [
        "ap.drop([\"Package\", \"Description\", \"Related apps\"],axis=1, inplace=True)"
      ],
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rb6AW7ccqbyr"
      },
      "source": [
        "cat = ap.Category.value_counts()\n",
        "cat = cat[cat>700]\n",
        "ap= ap[~ap[\"Category\"].isin(cat)]\n",
        "ap = pd.get_dummies(data=ap,prefix=\"\", columns=[\"Category\"], drop_first=True)"
      ],
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9CxWN-Csqlbt"
      },
      "source": [
        "ap = ap.drop_duplicates(subset=\"App\", keep=\"first\")"
      ],
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gSlGxZMtqpJS"
      },
      "source": [
        "ap = ap[~ap.App.str.contains(\"[?]\")]"
      ],
      "execution_count": 13,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Qxc8uBamqtdT",
        "outputId": "263aad76-64dc-4e65-ab67-510be7c10239"
      },
      "source": [
        "ap.App.value_counts()"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "SecretMail                        1\n",
              "Ragalahari                        1\n",
              "myFtWorth                         1\n",
              "City Of Mafia                     1\n",
              "Funny Videos and Pics by Break    1\n",
              "                                 ..\n",
              "Tennessee Titans Elite            1\n",
              "Troll Face Game                   1\n",
              "Dressup Maidens Avatar Deluxe     1\n",
              "Super Tether                      1\n",
              "Dice Job Search                   1\n",
              "Name: App, Length: 21420, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qydIU8fBqwmD"
      },
      "source": [
        "ap.set_index([\"App\"], inplace=True)"
      ],
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kwgf64UNqzNL",
        "outputId": "51b8114c-1a60-4fa8-d668-d5ee33d06831"
      },
      "source": [
        "ap.info()"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "Index: 21420 entries, Canada Post Corporation to Brain Ball Game\n",
            "Columns: 208 entries, Rating to _Weather\n",
            "dtypes: float64(3), int64(176), uint8(29)\n",
            "memory usage: 30.0+ MB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "s_SAKW3ysuOw"
      },
      "source": [
        "#Feature Selection"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aNutdQqmM4-S"
      },
      "source": [
        "X = ap.drop([\"Class\"],axis=1)\n",
        "y = ap[[\"Class\"]]\n",
        "\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.35, random_state=101)"
      ],
      "execution_count": 20,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TtxAWQaXM4-S"
      },
      "source": [
        "from sklearn.feature_selection import SelectKBest, mutual_info_classif\n",
        "fs = SelectKBest(score_func=mutual_info_classif, k=50)\n",
        "fs.fit(X_train, y_train)\n",
        "X_train_fs = fs.transform(X_train)\n",
        "X_test_fs = fs.transform(X_test)"
      ],
      "execution_count": 21,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 262
        },
        "id": "qM0JrBpAtNeS",
        "outputId": "f639848c-95e7-4e35-bbec-ff0bdffa43c5"
      },
      "source": [
        "plt.bar([i for i in range(len(fs.scores_))], fs.scores_)\n",
        "plt.show()"
      ],
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "EY5dY48btVI3",
        "outputId": "aea68de1-0b3f-4aa1-c76e-8536cbdbbc97"
      },
      "source": [
        "pd.DataFrame(data=zip(fs.scores_,fs.get_support()), index=X.columns.tolist(), columns=[\"score\", \"support\"]).sort_values(by=\"score\", ascending=False)[:51]"
      ],
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>score</th>\n",
              "      <th>support</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>Number of ratings</th>\n",
              "      <td>0.111195</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Price</th>\n",
              "      <td>0.092089</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Rating</th>\n",
              "      <td>0.074722</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>_Travel &amp; Local</th>\n",
              "      <td>0.024212</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>_Comics</th>\n",
              "      <td>0.022489</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Network communication : view network state (S)</th>\n",
              "      <td>0.016558</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Phone calls : read phone state and identity (D)</th>\n",
              "      <td>0.012594</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Your messages : write instant messages (D)</th>\n",
              "      <td>0.012123</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>_Tools</th>\n",
              "      <td>0.010779</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Safe permissions count</th>\n",
              "      <td>0.010400</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>_Libraries &amp; Demo</th>\n",
              "      <td>0.010100</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>_Education</th>\n",
              "      <td>0.009792</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Your personal information : read Browser's history and bookmarks (D)</th>\n",
              "      <td>0.009214</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Default : Advanced download manager functions. (S)</th>\n",
              "      <td>0.009172</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Dangerous permissions count</th>\n",
              "      <td>0.009041</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Your messages : edit SMS or MMS (D)</th>\n",
              "      <td>0.008906</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Your location : fine (GPS) location (D)</th>\n",
              "      <td>0.008877</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Your messages : receive WAP (D)</th>\n",
              "      <td>0.008682</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>_Cards &amp; Casino</th>\n",
              "      <td>0.008571</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Hardware controls : take pictures and videos (D)</th>\n",
              "      <td>0.008551</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Default : record what you type and actions you take (S)</th>\n",
              "      <td>0.007921</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Your accounts : Google Maps (D)</th>\n",
              "      <td>0.007883</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Your messages : Send Gmail (S)</th>\n",
              "      <td>0.007871</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Your accounts : Blogger (D)</th>\n",
              "      <td>0.007744</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Default : interact with a device admin (S)</th>\n",
              "      <td>0.007605</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>System tools : read sync settings (S)</th>\n",
              "      <td>0.007587</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>_News &amp; Magazines</th>\n",
              "      <td>0.007494</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>_Productivity</th>\n",
              "      <td>0.007481</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Your accounts : Picasa Web Albums (D)</th>\n",
              "      <td>0.007459</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>_Business</th>\n",
              "      <td>0.006953</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Network communication : download files without notification (S)</th>\n",
              "      <td>0.006810</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Default : read frame buffer (S)</th>\n",
              "      <td>0.006586</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Your personal information : add or modify calendar events and send email to guests (D)</th>\n",
              "      <td>0.006352</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Default : bind to an input method (S)</th>\n",
              "      <td>0.006185</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>System tools : set wallpaper (S)</th>\n",
              "      <td>0.005920</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Your accounts : contacts data in Google accounts (D)</th>\n",
              "      <td>0.005826</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>System tools : modify global animation speed (D)</th>\n",
              "      <td>0.005726</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Default : force application to close (S)</th>\n",
              "      <td>0.005675</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>System tools : allow Wi-Fi Multicast reception (D)</th>\n",
              "      <td>0.005307</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>_Casual</th>\n",
              "      <td>0.005253</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>_Racing</th>\n",
              "      <td>0.005182</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>_Music &amp; Audio</th>\n",
              "      <td>0.005179</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Network communication : Broadcast data messages to applications. (S)</th>\n",
              "      <td>0.005176</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Your accounts : Google Spreadsheets (D)</th>\n",
              "      <td>0.005144</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>_Medical</th>\n",
              "      <td>0.005103</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Your personal information : write Browser's history and bookmarks (D)</th>\n",
              "      <td>0.004966</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Storage : modify/delete USB storage contents modify/delete SD card contents (D)</th>\n",
              "      <td>0.004803</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Default : delete other applications' caches (S)</th>\n",
              "      <td>0.004758</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>System tools : change Wi-Fi state (D)</th>\n",
              "      <td>0.004505</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Your personal information : read contact data (D)</th>\n",
              "      <td>0.004364</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Your personal information : read sensitive log data (D)</th>\n",
              "      <td>0.004301</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                                       score  support\n",
              "Number of ratings                                   0.111195     True\n",
              "Price                                               0.092089     True\n",
              "Rating                                              0.074722     True\n",
              "_Travel & Local                                     0.024212     True\n",
              "_Comics                                             0.022489     True\n",
              "Network communication : view network state (S)      0.016558     True\n",
              "Phone calls : read phone state and identity (D)     0.012594     True\n",
              "Your messages : write instant messages (D)          0.012123     True\n",
              "_Tools                                              0.010779     True\n",
              "Safe permissions count                              0.010400     True\n",
              "_Libraries & Demo                                   0.010100     True\n",
              "_Education                                          0.009792     True\n",
              "Your personal information : read Browser's hist...  0.009214     True\n",
              "Default : Advanced download manager functions. (S)  0.009172     True\n",
              "Dangerous permissions count                         0.009041     True\n",
              "Your messages : edit SMS or MMS (D)                 0.008906     True\n",
              "Your location : fine (GPS) location (D)             0.008877     True\n",
              "Your messages : receive WAP (D)                     0.008682     True\n",
              "_Cards & Casino                                     0.008571     True\n",
              "Hardware controls : take pictures and videos (D)    0.008551     True\n",
              "Default : record what you type and actions you ...  0.007921     True\n",
              "Your accounts : Google Maps (D)                     0.007883     True\n",
              "Your messages : Send Gmail (S)                      0.007871     True\n",
              "Your accounts : Blogger (D)                         0.007744     True\n",
              "Default : interact with a device admin (S)          0.007605     True\n",
              "System tools : read sync settings (S)               0.007587     True\n",
              "_News & Magazines                                   0.007494     True\n",
              "_Productivity                                       0.007481     True\n",
              "Your accounts : Picasa Web Albums (D)               0.007459     True\n",
              "_Business                                           0.006953     True\n",
              "Network communication : download files without ...  0.006810     True\n",
              "Default : read frame buffer (S)                     0.006586     True\n",
              "Your personal information : add or modify calen...  0.006352     True\n",
              "Default : bind to an input method (S)               0.006185     True\n",
              "System tools : set wallpaper (S)                    0.005920     True\n",
              "Your accounts : contacts data in Google account...  0.005826     True\n",
              "System tools : modify global animation speed (D)    0.005726     True\n",
              "Default : force application to close (S)            0.005675     True\n",
              "System tools : allow Wi-Fi Multicast reception (D)  0.005307     True\n",
              "_Casual                                             0.005253     True\n",
              "_Racing                                             0.005182     True\n",
              "_Music & Audio                                      0.005179     True\n",
              "Network communication : Broadcast data messages...  0.005176     True\n",
              "Your accounts : Google Spreadsheets (D)             0.005144     True\n",
              "_Medical                                            0.005103     True\n",
              "Your personal information : write Browser's his...  0.004966     True\n",
              "Storage : modify/delete USB storage contents mo...  0.004803     True\n",
              "Default : delete other applications' caches (S)     0.004758     True\n",
              "System tools : change Wi-Fi state (D)               0.004505     True\n",
              "Your personal information : read contact data (D)   0.004364     True\n",
              "Your personal information : read sensitive log ...  0.004301    False"
            ]
          },
          "metadata": {},
          "execution_count": 23
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dyzHlUFgHvZ5"
      },
      "source": [
        "# Modelling "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GPN69-AGtZt4"
      },
      "source": [
        "feature_columns = [X.columns[i] for i in range(len(fs.get_support())) if fs.get_support()[i]]"
      ],
      "execution_count": 24,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "AHr_fdNktdCB",
        "outputId": "262aee4b-3726-40b5-9b27-2f72d11c184f"
      },
      "source": [
        "feature_columns"
      ],
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['Rating',\n",
              " 'Number of ratings',\n",
              " 'Price',\n",
              " 'Dangerous permissions count',\n",
              " 'Safe permissions count',\n",
              " 'Default : Advanced download manager functions. (S)',\n",
              " 'Default : bind to an input method (S)',\n",
              " \"Default : delete other applications' caches (S)\",\n",
              " 'Default : force application to close (S)',\n",
              " 'Default : interact with a device admin (S)',\n",
              " 'Default : read frame buffer (S)',\n",
              " 'Default : record what you type and actions you take (S)',\n",
              " 'Hardware controls : take pictures and videos (D)',\n",
              " 'Network communication : Broadcast data messages to applications. (S)',\n",
              " 'Network communication : download files without notification (S)',\n",
              " 'Network communication : view network state (S)',\n",
              " 'Phone calls : read phone state and identity (D)',\n",
              " 'Storage : modify/delete USB storage contents modify/delete SD card contents (D)',\n",
              " 'System tools : allow Wi-Fi Multicast reception (D)',\n",
              " 'System tools : change Wi-Fi state (D)',\n",
              " 'System tools : modify global animation speed (D)',\n",
              " 'System tools : read sync settings (S)',\n",
              " 'System tools : set wallpaper (S)',\n",
              " 'Your accounts : Blogger (D)',\n",
              " 'Your accounts : Google Maps (D)',\n",
              " 'Your accounts : Google Spreadsheets (D)',\n",
              " 'Your accounts : Picasa Web Albums (D)',\n",
              " 'Your accounts : contacts data in Google accounts (D)',\n",
              " 'Your location : fine (GPS) location (D)',\n",
              " 'Your messages : Send Gmail (S)',\n",
              " 'Your messages : edit SMS or MMS (D)',\n",
              " 'Your messages : receive WAP (D)',\n",
              " 'Your messages : write instant messages (D)',\n",
              " 'Your personal information : add or modify calendar events and send email to guests (D)',\n",
              " \"Your personal information : read Browser's history and bookmarks (D)\",\n",
              " 'Your personal information : read contact data (D)',\n",
              " \"Your personal information : write Browser's history and bookmarks (D)\",\n",
              " '_Business',\n",
              " '_Cards & Casino',\n",
              " '_Casual',\n",
              " '_Comics',\n",
              " '_Education',\n",
              " '_Libraries & Demo',\n",
              " '_Medical',\n",
              " '_Music & Audio',\n",
              " '_News & Magazines',\n",
              " '_Productivity',\n",
              " '_Racing',\n",
              " '_Tools',\n",
              " '_Travel & Local']"
            ]
          },
          "metadata": {},
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3cDIORGitfR8",
        "outputId": "bd781b6c-703d-46be-9bc0-1f911910d826"
      },
      "source": [
        "len(X_test_fs[0])"
      ],
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "50"
            ]
          },
          "metadata": {},
          "execution_count": 26
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SEbv6_Tn3ygA"
      },
      "source": [
        "# 1. Naive Bayes "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZG7EQZYo16jO"
      },
      "source": [
        "# train a Gaussian Naive Bayes classifier on the training set\n",
        "from sklearn.naive_bayes import GaussianNB\n",
        "\n",
        "# instantiate the model\n",
        "nb = GaussianNB()\n",
        "\n",
        "# fit the model\n",
        "nb.fit(X_train_fs, y_train)\n",
        "\n",
        "#predict the result\n",
        "pred_nb_train = nb.predict(X_train_fs)\n",
        "pred_nb_test = nb.predict(X_test_fs)"
      ],
      "execution_count": 33,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VQWVYxH24wTT",
        "outputId": "81499954-d201-4e5d-ca17-556b18a32af2"
      },
      "source": [
        "print(f\"Accuracy Score on train data: {metrics.accuracy_score(y_train, pred_nb_train)}\")\n",
        "print(f\"Accuracy Score on test data: {metrics.accuracy_score(y_test, pred_nb_test)}\")"
      ],
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy Score on train data: 0.5918264741794154\n",
            "Accuracy Score on test data: 0.5834333733493398\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "P59NrQKY2Rxm",
        "outputId": "ddfc0ae9-167e-4018-84f1-aecf6392526c"
      },
      "source": [
        "# Print the Confusion Matrix and slice it into four pieces\n",
        "\n",
        "from sklearn.metrics import confusion_matrix\n",
        "\n",
        "cm = confusion_matrix(y_test, pred_nb_test)\n",
        "\n",
        "print('Confusion matrix\\n\\n', cm)\n",
        "\n",
        "print('\\nTrue Positives(TP) = ', cm[0,0])\n",
        "\n",
        "print('\\nTrue Negatives(TN) = ', cm[1,1])\n",
        "\n",
        "print('\\nFalse Positives(FP) = ', cm[0,1])\n",
        "\n",
        "print('\\nFalse Negatives(FN) = ', cm[1,0])"
      ],
      "execution_count": 35,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Confusion matrix\n",
            "\n",
            " [[2415   57]\n",
            " [3066 1959]]\n",
            "\n",
            "True Positives(TP) =  2415\n",
            "\n",
            "True Negatives(TN) =  1959\n",
            "\n",
            "False Positives(FP) =  57\n",
            "\n",
            "False Negatives(FN) =  3066\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 299
        },
        "id": "51GrOk072aAx",
        "outputId": "1dac96a5-8c9f-44e8-9bdb-b90887fd1568"
      },
      "source": [
        "# visualize confusion matrix with seaborn heatmap\n",
        "\n",
        "cm_matrix = pd.DataFrame(data=cm, columns=['Actual Positive:1', 'Actual Negative:0'], \n",
        "                                 index=['Predict Positive:1', 'Predict Negative:0'])\n",
        "\n",
        "sns.heatmap(cm_matrix, annot=True, fmt='d', cmap='YlGnBu')"
      ],
      "execution_count": 36,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f63b9020110>"
            ]
          },
          "metadata": {},
          "execution_count": 36
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cs4Qi0f77ILO"
      },
      "source": [
        "#2. Random Forest"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XkCjouN37KBL"
      },
      "source": [
        "#create an instance of Select from Model. Pass an object of Random Forest Classifier with n_estimators=100 as argument. \n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "\n",
        "rf = RandomForestClassifier(n_estimators = 10, random_state = 30)\n",
        "\n",
        "#Fit the test\n",
        "rf.fit(X_train_fs, y_train)\n",
        "\n",
        "#predict the result\n",
        "pred_rf_train = rf.predict(X_train_fs)\n",
        "pred_rf_test = rf.predict(X_test_fs)"
      ],
      "execution_count": 37,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YaEfIu3x41Ga",
        "outputId": "70b37b1b-426d-445e-e790-8d32be5de12d"
      },
      "source": [
        "print(f\"Accuracy Score on train data: {metrics.accuracy_score(y_train, pred_rf_train)}\")\n",
        "print(f\"Accuracy Score on test data: {metrics.accuracy_score(y_test, pred_rf_test)}\")"
      ],
      "execution_count": 38,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy Score on train data: 0.9812540400775694\n",
            "Accuracy Score on test data: 0.7995198079231692\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nFojkgyJ_npm",
        "outputId": "73f7c751-aea3-468c-cd71-c9db43225044"
      },
      "source": [
        "# Print the Confusion Matrix and slice it into four pieces\n",
        "\n",
        "from sklearn.metrics import confusion_matrix\n",
        "\n",
        "cm = confusion_matrix(y_test, pred_rf_test)\n",
        "\n",
        "print('Confusion matrix\\n\\n', cm)\n",
        "\n",
        "print('\\nTrue Positives(TP) = ', cm[0,0])\n",
        "\n",
        "print('\\nTrue Negatives(TN) = ', cm[1,1])\n",
        "\n",
        "print('\\nFalse Positives(FP) = ', cm[0,1])\n",
        "\n",
        "print('\\nFalse Negatives(FN) = ', cm[1,0])"
      ],
      "execution_count": 39,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Confusion matrix\n",
            "\n",
            " [[1858  614]\n",
            " [ 889 4136]]\n",
            "\n",
            "True Positives(TP) =  1858\n",
            "\n",
            "True Negatives(TN) =  4136\n",
            "\n",
            "False Positives(FP) =  614\n",
            "\n",
            "False Negatives(FN) =  889\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 299
        },
        "id": "0Ny13ec__ohY",
        "outputId": "c616c14a-4e12-47e4-ad42-5e4f2a348842"
      },
      "source": [
        "# visualize confusion matrix with seaborn heatmap\n",
        "\n",
        "cm_matrix = pd.DataFrame(data=cm, columns=['Actual Positive:1', 'Actual Negative:0'], \n",
        "                                 index=['Predict Positive:1', 'Predict Negative:0'])\n",
        "\n",
        "sns.heatmap(cm_matrix, annot=True, fmt='d', cmap='YlGnBu')"
      ],
      "execution_count": 40,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f63b8c9f810>"
            ]
          },
          "metadata": {},
          "execution_count": 40
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Wm-4A3qHATwp"
      },
      "source": [
        "# 3. Decision Tree"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "y_h6wwyGAcAN"
      },
      "source": [
        "from sklearn.tree import DecisionTreeClassifier\n",
        "\n",
        "# Make a decision tree and train\n",
        "dt = DecisionTreeClassifier(criterion='entropy',max_depth =2,random_state=0)\n",
        "dt.fit(X_train_fs, y_train)\n",
        "\n",
        "#predict the result\n",
        "pred_dt_train = dt.predict(X_train_fs)\n",
        "pred_dt_test = dt.predict(X_test_fs)"
      ],
      "execution_count": 41,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PUrP_RRBv3_a",
        "outputId": "1f562827-4f33-490f-d28e-d70cbe8e9192"
      },
      "source": [
        "print(f\"Accuracy Score on train data: {metrics.accuracy_score(y_train, pred_dt_train)}\")\n",
        "print(f\"Accuracy Score on test data: {metrics.accuracy_score(y_test, pred_dt_test)}\")"
      ],
      "execution_count": 42,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy Score on train data: 0.7236227824463118\n",
            "Accuracy Score on test data: 0.7237561691343204\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "B4pgn4heYgBA",
        "outputId": "9a5fe20c-e7a1-4601-eee6-79af16fcaf7d"
      },
      "source": [
        "# Print the Confusion Matrix and slice it into four pieces\n",
        "\n",
        "from sklearn.metrics import confusion_matrix\n",
        "\n",
        "cm = confusion_matrix(y_test, pred_dt_test)\n",
        "\n",
        "print('Confusion matrix\\n\\n', cm)\n",
        "\n",
        "print('\\nTrue Positives(TP) = ', cm[0,0])\n",
        "\n",
        "print('\\nTrue Negatives(TN) = ', cm[1,1])\n",
        "\n",
        "print('\\nFalse Positives(FP) = ', cm[0,1])\n",
        "\n",
        "print('\\nFalse Negatives(FN) = ', cm[1,0])"
      ],
      "execution_count": 43,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Confusion matrix\n",
            "\n",
            " [[2019  453]\n",
            " [1618 3407]]\n",
            "\n",
            "True Positives(TP) =  2019\n",
            "\n",
            "True Negatives(TN) =  3407\n",
            "\n",
            "False Positives(FP) =  453\n",
            "\n",
            "False Negatives(FN) =  1618\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 299
        },
        "id": "Id6R3hRJYlGv",
        "outputId": "13860031-a45e-492d-bbcc-4d5a3c35e1d6"
      },
      "source": [
        "# visualize confusion matrix with seaborn heatmap\n",
        "\n",
        "cm_matrix = pd.DataFrame(data=cm, columns=['Actual Positive:1', 'Actual Negative:0'], \n",
        "                                 index=['Predict Positive:1', 'Predict Negative:0'])\n",
        "\n",
        "sns.heatmap(cm_matrix, annot=True, fmt='d', cmap='YlGnBu')"
      ],
      "execution_count": 44,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f63b9308990>"
            ]
          },
          "metadata": {},
          "execution_count": 44
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 592
        },
        "id": "0ZCVUXnqWZK9",
        "outputId": "2ed685b5-6e93-465f-b2e7-ebeb0f6eb911"
      },
      "source": [
        "fig=dt.fit(X_train_fs, y_train)\n",
        "plt.figure(figsize=(10,10))\n",
        "tree.plot_tree(fig)\n",
        "plt.show"
      ],
      "execution_count": 46,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<function matplotlib.pyplot.show>"
            ]
          },
          "metadata": {},
          "execution_count": 46
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FfdNQHlwbmyN"
      },
      "source": [
        "# 4. AdaBoost"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GA5dOnvFhrVZ"
      },
      "source": [
        "# Create adaboost classifer object\n",
        "ab = AdaBoostClassifier(n_estimators=30,\n",
        "                         learning_rate=1)\n",
        "# Train Adaboost Classifer\n",
        "ab.fit(X_train_fs, y_train)\n",
        "\n",
        "#predict the result\n",
        "pred_ab_train = ab.predict(X_train_fs)\n",
        "pred_ab_test = ab.predict(X_test_fs)"
      ],
      "execution_count": 47,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-xuriBq58NBa",
        "outputId": "8ff59986-9a92-4a11-e035-ba1678725874"
      },
      "source": [
        "print(f\"Accuracy Score on train data: {metrics.accuracy_score(y_train, pred_ab_train)}\")\n",
        "print(f\"Accuracy Score on test data: {metrics.accuracy_score(y_test, pred_ab_test)}\")"
      ],
      "execution_count": 48,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy Score on train data: 0.8107448107448108\n",
            "Accuracy Score on test data: 0.8153928237961852\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VsF8IrW3iLKd",
        "outputId": "1e516578-015b-4d8d-ef77-c071106c1ea2"
      },
      "source": [
        "# Print the Confusion Matrix and slice it into four pieces\n",
        "\n",
        "from sklearn.metrics import confusion_matrix\n",
        "\n",
        "cm = confusion_matrix(y_test, pred_ab_test)\n",
        "\n",
        "print('Confusion matrix\\n\\n', cm)\n",
        "\n",
        "print('\\nTrue Positives(TP) = ', cm[0,0])\n",
        "\n",
        "print('\\nTrue Negatives(TN) = ', cm[1,1])\n",
        "\n",
        "print('\\nFalse Positives(FP) = ', cm[0,1])\n",
        "\n",
        "print('\\nFalse Negatives(FN) = ', cm[1,0])"
      ],
      "execution_count": 49,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Confusion matrix\n",
            "\n",
            " [[1811  661]\n",
            " [ 723 4302]]\n",
            "\n",
            "True Positives(TP) =  1811\n",
            "\n",
            "True Negatives(TN) =  4302\n",
            "\n",
            "False Positives(FP) =  661\n",
            "\n",
            "False Negatives(FN) =  723\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 299
        },
        "id": "LVQMnJ3jiGCQ",
        "outputId": "a7f9ffa5-f5a9-4c4a-dc50-367ea2a34565"
      },
      "source": [
        "# visualize confusion matrix with seaborn heatmap\n",
        "\n",
        "cm_matrix = pd.DataFrame(data=cm, columns=['Actual Positive:1', 'Actual Negative:0'], \n",
        "                                 index=['Predict Positive:1', 'Predict Negative:0'])\n",
        "\n",
        "sns.heatmap(cm_matrix, annot=True, fmt='d', cmap='YlGnBu')"
      ],
      "execution_count": 50,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f63b9682410>"
            ]
          },
          "metadata": {},
          "execution_count": 50
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEJCAYAAACZjSCSAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3dd5xU1fnH8c8sSwcF7NIVeRCjiBqVGHsJItYYS9SoUdSIQSSJETVqLIixYIkNhYhGAVsSRH6iQWOJDbGgoI8gIIIGLCgiSp3fH/csDrg7e5ed2Z07833ndV8798y9d56Jwzxzyj0nlU6nERGR0lVW3wGIiEj9UiIQESlxSgQiIiVOiUBEpMQpEYiIlDglAhGREleej4veN3OixqRKzp3UpXN9hyBFqWsqV1dq2uH42N99384dnbPXra28JAIRkVKUSiWzkUWJQEQkR1IJbW1XIhARyRHVCERESlxZWYP6DmG9KBGIiOSMagQiIiVNTUMiIiVOiUBEpMRp1JCISIlTjUBEpMSVlSXzKzWZUYuIFKAUBTNrRI0oEYiI5IiahkRESpwSgYhIictFIjCzBsBrwHx372tmnYExwEbAFOAkd19uZo2Be4Gdgc+BY919TrjGYOA0YBUwwN0nZnvNZKYvEZGCVFaDrUrnAu9m7F8DDHP3LsAioi94wt9FoXxYOA4z6w4cB2wH9AZuC8kla9QiIpIDZWXlsbfKmFk74BDg7rCfAvYDHg6HjAKOCI8PD/uE5/cPxx8OjHH3Ze4+G5gJ7Jo17vV+xyIispYUZbG3KtwInA+sDvsbAV+6+8qwPw9oGx63BT4CCM9/FY5fU17JOZVSIhARyZFUqiz2ti4z6wssdPcpdR23EoGISI6kUqnYWyX2AA4zszlEncP7ATcBrcysoi2pHTA/PJ4PtAcIz29I1Gm8pryScyqlRCAikiO1qRG4+2B3b+funYg6e5929xOAZ4Cjw2EnA/8Kj8eFfcLzT7t7OpQfZ2aNw4ijbYBXs8WtRCAikiNlqfLYWw38ERhkZjOJ+gBGhPIRwEahfBBwAYC7TwMeBKYDTwD93X1VthdIpdPpGr3ROO6bOTH3F5WSd1KXzvUdghSlrjmbF6Jzz7/E/u6b/cb5BTMfhW4oExHJEU1DLSJS6jTFhIhIadNcQyIiJa6KYaEFT4lARCRHajgaqGAkM2oRkUKkGoGISIlLZheBEoGISM6oRiAiUuKUCERESlu6gRKBiEhpS2YeUCIQEcmZsmRmAiUCEZFcUR+BiEiJS2YeUCIQEckZNQ2JiJQ4JQIRkRKnRCAiUuKSmQeUCEREciWtUUMiIiVOTUMiIiUumXlAiUBEJGcarP881GbWBHgOaEz03fywu19qZvcAewNfhUNPcfc3zSwF3AT0AZaG8tfDtU4GLg7HX+nuo7K9thKBiEiu1K5GsAzYz92XmFlD4AUz+7/w3B/c/eF1jj8Y2CZsuwG3A7uZWRvgUmAXIA1MMbNx7r6oqhdWIhARyZVadBa7expYEnYbhi2d5ZTDgXvDeS+bWSsz2wLYB3jK3b8AMLOngN7A6KoulND1dEREClAqFX+rhJk1MLM3gYVEX+avhKeuMrOpZjbMzBqHsrbARxmnzwtlVZVXSYlARCRXymqwVcLdV7n7jkA7YFcz+xEwGOgG/BhoA/wxH2GLiEgulKXib1m4+5fAM0Bvd//E3dPuvgz4G7BrOGw+0D7jtHahrKryKqmPIMceu/F+Zrw6jeatWnLmbYMB+N8H8/i/W8eycvlKyhqU0fvsY2hrHfnsowU8duP9/G/mR+zzq770+vn+Wa8jUmHx4iVcfPEtvP/+h6RSKYYMOZeePbtx332Pcf/9j9OgQRl77/1jzj//VBYtWsyAAUN5550ZHHnk/lxyyVn1HX7RStfiPgIz2wRY4e5fmllT4EDgGjPbwt0/CaOEjgDeCaeMA84xszFEncVfheMmAkPMrHU47iCiWkWVlAhybIcDdmOXvnsx7oa/rymb9Ld/secvD6bLLt2ZOXkak/72L341dABNWzbjZ2f+HH/p7VjXEalw1VV3seeeO3HzzYNZvnwF3323jJdfnsqkSa8wbtwtNGrUkM8//xKAxo0bce65JzBjxlxmzPiwniMvcrW7s3gLYJSZNSBqrXnQ3ceb2dMhSaSAN4GKTD6BaOjoTKLho6cCuPsXZnYFMDkcd3lFx3FV1isRmNn/ufvB63Nusev4oy58ueDztcpSqRTLln4HwHfffEfLNhsC0LxVS5q3asmMydNiXUcE4Ouvv2Hy5HcYOnQgAI0aNaRRo4aMHj2BM844mkaNGgKw0UatAGjWrAm77LIdc+d+Um8xl4xa5AF3nwr0rKR8vyqOTwP9q3huJDAy7mtXmQjMbKcqnkoBO8Z9AYGD+h3FA5fczr9H/BPSaU6+7rz6DkkSbN68BbRpsyGDB9/Ie+/NYbvttuaii85gzpyPee21aQwbdh+NGzfk/PN/zQ47dK3vcEtLEU4xMRl4lspzXKv8hFOcpkx4gQP7Hcm2e+zI9OdfZ/yND3DikHPqOyxJqJUrVzF9+gf86U9n0qOHceWVwxk+/GFWrVrFV18t4cEHr+Ptt2cwcOA1TJp0N6mEToSWSAn9/zpbIngXONPdZ6z7hJl9VMnxUoWpk17loDN/DsC2P+3J+JuqvK9DpFqbb74xm2++MT16GAC9e+/B8OEPs9lmG3Pggb1IpVLssENXysrKWLRoMW1CU6TUgQbJTATZho9eluX53+Y+lOLVos2GfPj2TADmvPU+bbbcpJ4jkiTbZJPWbL75xsyaNQ+Al156i623bs8BB+zOK69MBWD27PmsWLGS1q03qM9QS0+Oho/WtVQ6ne0O5vVz38yJub9oQjx6zT3MfXsmSxcvoXmrlux1Qh82arcpT975CKtXr6a8YUMOPvsXbLFNB5Z8sZgRA69l2dLvSJWV0ahJI86640IaN2ta6XV6/qxXfb+9enVSl871HULBePfdWVx00S2sWLGS9u034+qrB9K0aWMuvPBm3ntvFg0blnP++b+mV68eAOy332ksWbKUFStW0rJlc0aOvJwuXTrU87soFF1z9q281ekPxf7um3X3LwomG8RKBGa2U8WsdpXtr6uUE4HkjxKB5EcOE8EZD8dPBMOPLphEEPfO4t9Usy8iIrWca6i+xLqPwN37ZdsXEREKru0/rmoTQbit+QRgK3e/3Mw6AJu7+6t5j05EJEmKcNRQhduAXsDxYf9r4Na8RSQiklQJHTUUJxHs5u79ge8Awio3jfIalYhIAqVTqdhbIYnTR7AiTIKUhjUz5K3Oa1QiIkmU0In94ySCm4F/AJua2VXA0Xy/KLKIiFQosCafuKpNBO5+v5lNAfYnmnfoCHd/N++RiYgkTYNkVgnijBq6GRjj7uogFhHJJpkVglhNQ1OAi83MiJqIxrj7a/kNS0QkeWqzQll9qrYe4+6j3L0P0cLJTrR02g9mJBURKXkJHT5akxXKugDdgI5EU1SLiEimAhsWGlecPoK/AEcCHwBjgSvc/ct8ByYikjjJ7CuOVSP4AOjl7p/lOxgRkUQrtlFDZtbN3d8jWrKyQ5hjaI1s01CLiJSkAmv7jytbjWAQcAZwfSXPpYH98hKRiEhC1WbqCDNrAjwHNCb6bn7Y3S81s87AGGAjolGcJ7n7cjNrDNwL7Ax8Dhzr7nPCtQYDpwGrgAHuPjHba1dZj3H3M8LDg91938wN6LPe71ZEpFiV1WD7oWXAfu7eA9gR6G1muwPXAMPcvQuwiOgLnvB3USgfFo7DzLoDxwHbAb2B28I0QVnDrs6LMctEREpbLRamcfe0uy8Juw3DVtH68nAoHwUcER4fHvYJz+8flg04nOh+r2XuPhuYCeyaLexsfQSbA22BpmbWk+/vmdsAaJbtoiIiJamWfQThl/sUouH6txIN1vnS3VeGQ+YRfS8T/n4E4O4rzewrouajtsDLGZfNPKdS2foIfgacArQDbsgo/xq4sNp3JCJSamq5MI27rwJ2NLNWRDM5dMtFWNWpMhG4+yhglJn93N0fqYtgRESSLFdTTLj7l2b2DNGiYK3MrDzUCtoB88Nh84H2wDwzKwc2JOo0riivkHlOpbI1DZ3o7n8HOpnZoEoCvaGS00RESlftRg1tAqwISaApcCBRB/AzRNP/jwFOBv4VThkX9l8Kzz/t7mkzGwc8YGY3AFsC2wBZlxbO1jTUPPxtsV7vSkSk1NSuRrAFUStMA6KBPA+6+3gzmw6MMbMrgTeAEeH4EcB9ZjYT+IJopBDuPs3MHgSmAyuB/qHJqUqpdDpdm8Ardd/Mibm/qJS8k7p0ru8QpCh1zdldYB1ufjb2d9/cAXsXzN1ncecauhL4FngC2AE4LzQbiYhI0CDraP3CFec+goPcfTHQF5hDNKzpD/kMSkQkiWpxG0G9ipMIKmoNhwAPuftXeYxHRCSxUqlU7K2QxJl9dLyZvUfUNPSb0LP9XX7DEhFJngL7fo8tzgplFwA/AXZx9xXAN0S3MIuISIakNg3F6SxuCJwI7BUtW8yzwB15jktEJHFSyVyOIFbT0O1Ekx/dFvZPCmWn5ysoEZEkSui6NLESwY/DtKgVnjazt/IVkIhIUhVak09ccfLXKjPbumLHzLYiWuxAREQyFG0fAdE9A8+Y2Syiqag7AqfmNSoRkQQqtGGhcWVNBGGo6FdEixpsGord3ZflOzARkaQpus5iMzsdGEK0MEJn4Ax3H1dXgYmIJE1ZQhNBtrAHAtu5ey+i+wgG101IIiLJlNQ+gmyJYLm7fwrg7rOAxnUTkohIMpWl4m+FJFsfQTszu7mqfXcfkL+wRESSp9B+6ceVLRGsO8PolHwGIiKSdEWXCMKaxSIiElOq0Np8YopzH4GIiMSQ1FFDSgQiIjmS1KahavOXme0Rp0xEpNQlddRQnIrMLTHLRERKWlLvI8h2Z3HFjWSbmNmgjKc2ABK6RLOISP7UZooJM2sP3AtsBqSB4e5+k5ldBvQDPg2HXujuE8I5g4HTiCYCHeDuE0N5b+Amou/qu919aLbXztZH0AhoEY5pmVG+GDi6Jm9QRKQU1PKX/krgd+7+upm1BKaY2VPhuWHufl3mwWbWHTgO2A7YEvi3mXUNT98KHAjMAyab2Th3n17VC2cbPvos8KyZ3ePuH67vOxMRKRVltWj8d/dPgE/C46/N7F2gbZZTDgfGhElAZ5vZTKIJQgFmhhkhMLMx4dgqE0GciszdZtaqYsfMWpvZxBjniYiUlFz1EZhZJ6An8EooOsfMpprZSDNrHcraAh9lnDYvlFVVXqU4w0c3dvcvK3bcfZGZbZrthBO2bhfjsiI107TDpfUdghShb+eOztm1ctEJbGYtgEeAge6+2MxuB64g6je4Arge+HXtX+l7cRLBajPr4O5zQ5AdQ0AiIpKhtsNCzawhURK4390fBXD3BRnP3wWMD7vzgfYZp7cLZWQpr1ScRHAR8IKZPUu0QtmewBkxzhMRKSm1SQRmlgJGAO+6+w0Z5VuE/gOAI4F3wuNxwANmdgNRZ/E2wKtE39PbmFlnogRwHPDLbK9dbSJw9yfMbCdg91A00N0/i/vmRERKRXlZrRpL9gBOAt42szdD2YXA8Wa2I1FLzBzgTAB3n2ZmDxJ1Aq8E+rv7KgAzOweYSDR8dKS7T8v2wql0uvLAzaybu78XksAPuPvrVV10dXqamo4k55p3vLK+Q5Ai9O3c0Tm7veuQJ1+I/d33+EE/LZjbyrLVCH5HdBPD9ZU8lwb2y0tEIiIJVZZK5m/gbPcR9At/9627cEREkqvQ5hCKK9sUE0dlO7GiR1tERCIJnYU6a9PQoeHvpkRzDj0d9vcFXgSUCEREMhRdjcDdTwUwsyeB7hXDl8xsC+CeOolORCRBGtRu1FC9iXMfQfuMMawAC4AOeYpHRCSxirFpqMKkMLdQxX3YxwL/zl9IIiLJlNRRQ9UmMHc/B7gD6BG24e7+23wHJiKSNMW8QhnA68Dj7n4eMDHMlS0iIhnKarAVkjhrFvcDHgbuDEVtgX/mMygRkSQq5hpBf6I5MBYDuPsMoiGlIiKSobwsHXsrJHESwTJ3X16xY2blaBpqEZEfSGrTUJxRQ8+a2YVAUzM7EDgbeCy/YYmIJE/RjhoC/gh8CrxNNP3pBODifAYlIpJESe0jyFojMLMGwDR37wbcVTchiYgkU6F9wceVtUYQFjlwM9OdxCIi1ShPpWNvhSROH0FrYJqZvQp8U1Ho7oflLSoRkQRKao0gTiL4U96jEBEpAoU2GiiubOsRNAHOAroQdRSPcPeVdRWYiEjSFGONYBSwAngeOBjoDpxbF0GJiCRRqsDa/uPKlgi6u/v2AGY2Ani1bkISEUmmYqwRrKh44O4rzawOwhERSa7ajAYys/bAvcBmRLM3DHf3m8ysDTAW6ATMAY5x90VmlgJuAvoAS4FT3P31cK2T+f5+ryvdfVS2187Wt9HDzBaH7Wtgh4rHZrZ4fd+siEixquUNZSuB37l7d2B3oL+ZdQcuACa5+zbApLAPUZP9NmE7A7gdICSOS4HdgF2BS82sdba4sy1V2SD2uxcRkVo1DYWVID8Jj782s3eJZns+HNgnHDYK+A/RjA+HA/e6exp42cxahaWE9wGecvcvAMzsKaA33y8u9gNxho+KiEgMufr1bGadgJ7AK8BmGcsF/4+o6QiiJPFRxmnzQllV5VVK6rBXEZGCU5ZKx96qYmYtgEeAge6+VjN8+PWf86FJSgQiIjlS20nnzKwhURK4390fDcULQpMP4e/CUD4faJ9xertQVlV51XHX7G2KiEhVGpbF39YVRgGNAN519xsynhoHnBwenwz8K6P8V2aWMrPdga9CE9JE4CAzax06iQ8KZVVSH4GISI7U8j6CPYCTgLfN7M1QdiEwFHjQzE4DPgSOCc9NIBo6OpNo+OipAO7+hZldAUwOx11e0XFcFSUCEZEcqc3CNO7+AlBVKtm/kuPTREsJV3atkcDIuK+tRCAikiPFeGexiIjUQFJvvlIiEBHJkfKy4pt0TkREaqCBmoZEREqb+ghEREqcEoGISIlTIhARKXENinCFMhERqYFy1QhEREqbmoZEREqcmoZEREqcagQiIiVOiUBEpMQpEYiIlLiGmmtI1jV71nwGDbp+zf5HHy3gtwOOY+GCL3jmmddo2LCc9h02Y8iQ37LBBs2ZOnUGl15yOwDpdJr+5xzLgQfuXl/hSwEqK0vx3/FD+HjBF/z81Gu5/S9nsNMOW5FKpZg5+xP6Dbqdb5Yuo1GjckYMO5ue23fmi0VLOLH/Tcyd9xn77bk9V1xwHI0alrN8xUouvOoBnn1xWn2/raKR1CUfU+l07jPY6vS0ZKbFPFq1ahX77N2PMWOHMmf2x+y2+/aUlzfguuvuBeD3v/8V3367jIYNyykvb8DChV9w5BGDePa5EZSXJ3Vy29xq3vHK+g6h3g04vQ877bAVLVs25eenXkvLFk35esm3AFzzpxP59PPFXHfbOM446UB+tG0HBlw4gl8c2ovDev+Yk/rfTI/tOrHws6/4ZMEiundtx2N/H8zWu1a6tknJ+Hbu6Jw16Dz98YTY3337bdmnYBqSstYIwhqauwJtQ9F84NWwMo7UwMsvvU379pvRtu2mtG276ZryHj268uTElwBo2rTxmvLly1eQShXM50QKQNvN29B7/55cc8s/GdCvD8CaJADQpEkjKn7Y9T1oZ64a9ggAj054hRuuOBWAt6bNWXP89Pfn0aRJIxo1Kmf58pV19C6KW9HNPmpmBwG3ATOIEgBAO6CLmZ3t7k/WQXxFY8KEFzjkkD1/UP7oI09zcJ891uy/9db7XHTRrXzy8acMvWaAagOyxrWX/YqLhjxAi+ZN1iq/87oz+dm+PXlvxjwuuOLvAGy5eRvmffw5AKtWrWbx10vZqHVLPl/09ZrzjuyzK2++M1tJIIdqs1RlfcpWI7gJOMDd52QWmllnokWTt81jXEVl+fIVPP30ZM4bdOJa5Xfc8TANyss49NC91pT16NGV8eNv4oMP5jH4gpvZa6+daNy4UV2HLAXm4P17svCzxbzx9mz23H3tf3pn/v5OyspS3HD5qRx9aC/ue+jZaq+3bdd2XDn4l/Q9cUi+Qi5J5QntJMgWdjkwr5Ly+UDD/IRTnJ5//g26d9+KjTdutabsH48+zX+eeY1rrz2v0iagrbduR7NmTZjx/ty6DFUKVK9djL4H7sR7/72Ze/86gH1+sh0jb/y+bX/16jQPjXuRI/rsCsDH//uCdltuBECDBmVs0LLZmtpA283bMHb4IE4/7zZmf7iw7t9MESurwVZIstUIRgKTzWwM8FEoaw8cB4zId2DF5PHHn+eQQ366Zv/5519nxIh/cu99V6zVLzBv3gI233xjyssbMH/+QmbNmk/bdptWdkkpMZdcM4ZLrhkDwJ67b8vAM/vy64G3slXHzZj14QIA+h64M+/P/BiAx5+awglH78Urr8/gqD67rRkZtOEGzXj0nvP509DRvPTa+/XzZopYbbv1zGwk0BdY6O4/CmWXAf2AT8NhF7r7hPDcYOA0YBUwwN0nhvLeRK06DYC73X1ottetMhG4+9Vm9k/gcKBXKJ4PnODu09fnTZaipUu/48X/vsWf/3zWmrIrr7ib5ctXcNqv/wxEzUGX/fkspkx5l7vu+gcNyxuQKktxyaVn0Lr1BvUVuhS4VCrF3cN+Q8sWTUmlUrw9/UMGXDQSgHvG/oeRN57NO88NY9GXSzjpnFsAOOvkn7F1p80YfO5RDD73KAAOPfFqPv18cb29j2KSg77ie4C/AveuUz7M3a/LLDCz7kQ/zLcDtgT+bWZdw9O3AgcStepMNrNx2b63NXxUEkPDRyUfcjl89LXPHo/93bfLxodU+rpm1gkYv06NYEkliWAwRD/aw/5E4LLw9GXu/rPKjqtMrKaqEEiV+yIiktc+gnPMbKqZjTSz1qGsLd8320P0679tlvKscccxpZp9EZGSV5ZKx95q4HZga2BH4BPg+uyH11ysKSbc/bFs+yIiUvvO4sq4+4KKx2Z2FzA+7M4nGsBToR3f3/NVVXmlqq0RmFlXM5tkZu+E/R3M7OLqwxcRKS2pGmxxmdkWGbtHAu+Ex+OA48yscbi/axvgVWAysI2ZdTazRkQdyuOyvUacGsFdwB+AOwHcfaqZPQCo505EJENtp6E2s9HAPsDGZjYPuBTYx8x2BNLAHOBMAHefZmYPAtOBlUB/d18VrnMOMJFo+OhId886s2CcRNDM3V81s8wy3ZMuIrKO2rYMufvxlRRXed+Wu18FXFVJ+QSiGSBiiZMIPjOzrYmyEWZ2NFGHhYiIZCjmhWn6A8OBbmY2H5gNnJDXqEREEiiheSBWIvjQ3Q8ws+ZAmbt/Xe0ZIiIlKKk1gjj3Ecw2s+HA7sCSPMcjIpJY+Rg1VBfiJIJuwL+Jmohmm9lfzeyn1ZwjIlJyUql07K2QVNs05O5LgQeBB8OtzTcBzxINSxIRkaDQfunHFevOYjPbGzgW6A28BhyTz6BERJKo6JaqrGBmc4A3iGoFf3D3b/Ick4hIIiV1mfE4NYId3F2TlYuIVCOheSDr4vXnu/tfgKvM7Ac9G+4+IK+RiYgkTDHWCN4Nf1+ri0BERJIuoXkg61KVFVNNL3X3hzKfM7Nf5DUqEZEEKuYbygbHLBMRKWl5Wpgm77L1ERwM9AHamtnNGU9tgGYfFRH5gYRWCLL2EXxM1D9wGGsvTfk1cF4+gxIRSaKi6yx297eAt8zsAXdfUYcxiYgkUkLzQKz7CDqZ2dVAd6BJRaG7b5W3qEREEihOp2shihP334DbifoF9gXuBf6ez6BERJIolUrF3gpJnETQ1N0nASl3/9DdLwMOyW9YIiLJk6rB/wpJnKahZWZWBswICyLPB1rkNywRkeRJpZLZOBQnEZwLNAMGAFcA+wEn5zMoEZFkqt0vfTMbCfQFFrr7j0JZG2As0AmYAxzj7ovMLEW0LEAfYClwiru/Hs45Gbg4XPZKdx+V7XXjrEcwOTxcApxas7clIlI6ctDkcw/wV6K+2AoXAJPcfaiZXRD2/wgcDGwTtt2I+nJ3C4njUmAXIA1MMbNx7r6oqheNMw31Y+Fimb4iusfgTnf/LtbbExEperVLBO7+nJl1Wqf4cGCf8HgU8B+iRHA4cK+7p4GXzayVmW0Rjn3K3b8AMLOniNaSGV3V68Zp0JpFVBu4K2yLiW4q6xr2RUQESKUaxN5qYDN3/yQ8/h+wWXjcFvgo47h5oayq8irF6SP4ibv/OGP/MTOb7O4/NrNpMc4XESkJ+R4N5O7pypYFqK04NYIWZtahYic8rhg1tDzXAYmIJFWeho8uCE0+hL8LQ/l8oH3Gce1CWVXlVYpTI/gd8IKZfUDUANYZONvMmhO1V4mICJCne4vHEY3UHBr+/iuj/BwzG0PUWfyVu39iZhOBIWbWOhx3ENXMGB1n1NAEM9sG6PZ90ZoO4htr8m5ERIpZbe8YNrPRRJ29G5vZPKLRP0OBB83sNOBD4Jhw+ASioaMziYaPngrg7l+Y2RVAxYjPyys6jquMO53O3txkZs2AQUBHd+8XkoK5+/iqzlmdnlZYk21LUWje8cr6DkGK0LdzR+esYf+blc/F/u5rXr5XwdxeHHeuoeVAr7A/H9C/SBGRdaRoEHsrJHESwdZhEfsVAO6+lOTOtioikjfFPNfQcjNrSripzMy2BpblNSoRkQQqtFlF44qTCC4FngDam9n9wB7AKfkMSkQkmYo0Ebj7U2b2OrA70bs8190/y3tkIiIJk0ro0jTZFq/vsE7R2+FvMzPr4O5z8xeWiEjyFOM01I8T9Qtk1nXSwCbAplBg3d4iIvWuyJqG3H37zP0wI94fgQOAIfkNS0QkeYquaahCuIHsIqJbmK8HBrj7inwHJiKSPEVWIzCzHxElgO2AvwCnufuqugpMRCRpCu3+gLiy1QjeIprT+nFgV2BXM1vzpLsPyG9oIiLJUoz3Efy6zqIQESkChTZ1RFzZOos1xbSISI0UX41ARERqoBibhkREpEaSOXw0znoEe7j7f6srExGRZIqTvm6JWSYiIhgAsE4AAAgVSURBVAmU7T6CXsBPgE3MbFDGUxug6SVERIpGtj6CRkCLcEzLjPLFwNH5DEpEROpOnD6Cju7+YR3Fk3NmdgTwD2Bbd3+vmmMHAsPDKmzr81qnALu4+zmVlF9LtMxnI2CYu99Vw2ufBSx193vD9Z5094/Dc3cDN7j79PWJO+M1fgFcBmwL7Orur9XmesWogD5PI4Ed3X1qKHsH6Ovuc9bntap4/R2BLd19Qtg/DOju7kNzcO3BwGnAKqJpaybW9pqy/uL0EdxtZq0qdsystZkl6T/a8cAL4W91BgLN8hTHWHffEdgHGGJmm9XkZHe/w93vDbunAFtmPHd6bZNA8A5wFPBcDq5VrArl8zSPaAqYfNoR6FOx4+7jcpQEugPHEU1f0xu4zczU3FyP4gwf3djdv6zYcfdFZrZpHmPKGTNrAfwU2Bd4jGi1NcKH7hqiD+Fq4C6iO0G2BJ4xs8/cfV8zW+LuLcI5RxP94jrFzA4FLib6df85cIK7L4gTk7svNLMPgI5hPqfriP47TAZ+4+7LzGwocBiwkuiX/+/N7DJgCTAH2AW438y+BXoB/wf8PpRv7e5/CDGfQvhFaWYnAgNCzK8AZ687d5S7vxvOi/X/b6kpsM/TeGAvMzN393XiPAj4M9AY+AA41d2XmFkf4AbgG+C/wFbu3tfMdgVuApoA3wKnArOBy4GmZvZT4GqgKdFn7CJgKtDZ3VebWXPgPWAroANwK9F09UuBfpXUnA4Hxrj7MmC2mc0kmsbmpWres+RJnBrB6sxFasysI2H94gQ4HHjC3d8HPjeznUP5GUAnoqr1DsD97n4z8DGwr7vvW811XwB2d/eewBjg/LgBmdlWRP9g5gH3AMeGKb/Lgd+Y2UbAkcB2IbYrM89394eB14i+LHZ0928znn4knFvhWGCMmW0bHu8RaiWrgBNCPHeb2S5x4y9xhfR5Wk00GeSFmYVmtjFRUjnA3Xci+qwMMrMmwJ3Awe6+M9EXdYX3gD3D618CDHH35eHx2PA5G1txsLt/BbwJ7B2K+gITw6zEw4Hfhtf4PXBbiOswM7s8HN+WaB6zCvNCmdSTODWCi4AXzOxZol85exJ98JPgeKJfOhD9AzsemEK0psId7r4SwN2/qOF12wFjzWwLol9xs2Occ2z4ZbUMOJPoH+Ls8KUCMAroD/wV+A4YYWbjiX75xeLun5rZLDPbHZgBdCP65dcf2BmYHH7tNwUWhnNOj3t9KajPE8ADwEVm1jmjbHegO/Df8N+6EdEv7W7ALHevuPZovv93vCEwKkw5nwYaxnjtsUQ/Lp4haua5LdSYfgI8lFGrbAxRsxIwLub7kjoWZ83iJ8xsJ6IPGMDAJKxZbGZtgP2A7c0sTTTkNW1mf6jBZTJrPk0yHt9C1Dk7zsz2Iepgrc7YzE4/M+tR2UHuvjJU1fcnGp11TngfcY0BjiH6lfcPd0+bWQoY5e6Da3AdyVCAn6eKz8r1RAtGVUgBT7n7Wn0YoeO3KlcAz7j7kWEBqv/EePlxRH1dbYh+ZDwNNAe+DLXObOYD7TP224UyqSdVNg2ZWbfwdyeidr+Pw9YhlBW6o4H73L2ju3dy9/ZEv7T2BJ4CzjSzcljzjxzga9YeKrvAzLY1szLWbnLZkO8/uCevZ3wOdDKzLmH/JODZ8KtqwzBS4zygsoSxbpyZ/kHUhHE8UVIAmAQcXdG3Y2ZtQhOfxFeon6d7iGokFU09LwN7VHyuzKy5mXUl+rxtFb7oIfo1X9nrn5JRXuXnzN2XEPVr3QSMd/dV7r6YqM3/F+G1U1X84BkHHGdmjUNtZhvg1djvWHIuWx/B78Lf6yvZrstzXLlwPNGXYqZHQvndwFxgqpm9BfwyPD8ceMLMngn7FxA1zbwIfJJxncuIqr9TgPWqHbn7d0Sdcg+Z2dtEbb53EP3DG29mU4najgdVcvo9wB1m9qaZNV3nuouAd4GO7v5qKJtO1G78ZLjuU8AWsHYfgZkdaWbziDqgH0/Y6LB8K8jPU2jLv5loHXHc/VOiL/PR4b/1S0C30Jd0dohnCtGX/FfhMn8BrjazN1i7leAZoHv4nGUmjgpjgRPD3wonAKeF/x+mEf0oWauPwN2nAQ8C04EngP5a9Kp+VXsfgYgUBzNrEUYPpYhG9sxw92H1HZfUv2xTTByV7UR3fzT34YhIHvUzs5OJOpDfIBpFJFJ1jcDM/hYebko0EuDpsL8v8KK7981/eCIikm/ZVig7FcDMniS6rfyTsL8FURu1iIgUgTg3lLWvSALBAqJRRCIiUgTi3FA2KYweGR32jwX+nb+QRESkLsUaNWRmRwJ7hd3n3H3dYXQiIpJQcRfYfB143N3PAyaaWVU3M4mISMJUmwjMrB/wMN8PNWsL/DOfQYmISN2JUyPoD+xBtDIZ7j6DcBejiIgkX5xEsCzcxg5AmE9FtyOLiBSJOIngWTO7kGiBigOBh4gW5RARkSIQJxH8EfgUeJtoHv0JRBOYiYhIEcg6fDQswTfN3bvVXUgiIlKXstYIwtSwnrlUpYiIFJc4dxa3BqaZ2atEi14D4O6H5S0qERGpM3ESwZ/yHoWIiNSbbOsRNAHOAroQdRSPqFicW0REike2PoJRwC5ESeBgoiUqRUSkyGRrGuru7tsDmNkItLi0iEhRylYjWFHxQE1CIiLFK9tSlav4fpRQCmgKLA2P0+6+QZ1EKCIieRVrPQIRESlecdcjEBGRIqVEICJS4pQIRERKnBKBiEiJUyIQESlxSgQiIiXu/wEBydVIQCL4PQAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "B8jY87LbH1no"
      },
      "source": [
        "# Model Evaluation\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 388
        },
        "id": "GEP9HHYU-g9B",
        "outputId": "2675a8c7-f87f-48a1-8050-a2e6a7d49e8f"
      },
      "source": [
        "from sklearn.metrics import roc_curve\n",
        "from sklearn.metrics import roc_auc_score\n",
        "\n",
        "# Naive Bayes Classifier\n",
        "nb_clf = GaussianNB()\n",
        "nb_clf.fit(X_train_fs, y_train)\n",
        "nb_prediction_proba = nb_clf.predict_proba(X_test_fs)[:, 1]\n",
        "\n",
        "# Decision Tree Classifier\n",
        "dt_clf = DecisionTreeClassifier(max_depth =2,random_state=0)\n",
        "dt_clf.fit(X_train_fs, y_train)\n",
        "dt_prediction_proba = dt_clf.predict_proba(X_test_fs)[:, 1]\n",
        "\n",
        "# AdaBooster Classifier\n",
        "ab_clf = AdaBoostClassifier(n_estimators=50,learning_rate=1)\n",
        "ab_clf.fit(X_train_fs, y_train)\n",
        "ab_prediction_proba = ab_clf.predict_proba(X_test_fs)[:, 1]\n",
        "\n",
        "# Ranodm Forest Classifier\n",
        "rf_clf = RandomForestClassifier(n_estimators=20)\n",
        "rf_clf.fit(X_train_fs, y_train)\n",
        "rf_prediction_proba = rf_clf.predict_proba(X_test_fs)[:, 1]\n",
        "\n",
        "def roc_curve_and_score(y_test, pred_proba):\n",
        "    fpr, tpr, _ = roc_curve(y_test,pred_proba)\n",
        "    auc = roc_auc_score(y_test,pred_proba)\n",
        "    return fpr, tpr, auc\n",
        "plt.figure(figsize=(8, 6))\n",
        "plt.grid()\n",
        "\n",
        "fpr, tpr, roc_auc = roc_curve_and_score(y_test, rf_prediction_proba)\n",
        "plt.plot(fpr, tpr, color='darkorange', lw=2,\n",
        "         label='Random Forest={0:.3f}'.format(roc_auc))\n",
        "\n",
        "fpr, tpr, roc_auc = roc_curve_and_score(y_test, nb_prediction_proba)\n",
        "plt.plot(fpr, tpr, color='green', lw=2,\n",
        "         label='Naive Bayes={0:.3f}'.format(roc_auc))\n",
        "\n",
        "fpr, tpr, roc_auc = roc_curve_and_score(y_test, dt_prediction_proba)\n",
        "plt.plot(fpr, tpr, color='red', lw=2,\n",
        "         label='Decision Tree={0:.3f}'.format(roc_auc))\n",
        "\n",
        "fpr, tpr, roc_auc = roc_curve_and_score(y_test, ab_prediction_proba)\n",
        "plt.plot(fpr, tpr, color='blue', lw=2,\n",
        "         label='AdaBoost={0:.3f}'.format(roc_auc))\n",
        "\n",
        "plt.plot([0, 1], [0, 1], color='navy', lw=1, linestyle='--')\n",
        "plt.legend(loc=\"lower right\")\n",
        "plt.xlim([0.0, 1.0])\n",
        "plt.ylim([0.0, 1.05])\n",
        "plt.xlabel('1 - Specificity')\n",
        "plt.ylabel('Sensitivity')\n",
        "plt.show()"
      ],
      "execution_count": 57,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IdhwuQqLlNaw"
      },
      "source": [
        "#Saving Files"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Uwec_L3-jtEQ"
      },
      "source": [
        "import joblib\n",
        "with open(\"model.joblib\", 'wb') as model:\n",
        "    joblib.dump(rf, model, compress=1)"
      ],
      "execution_count": 58,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "n2cCMe1Smitz"
      },
      "source": [
        "with open(\"features.joblib\", 'wb') as ft:\n",
        "    joblib.dump(feature_columns, ft, compress=1)"
      ],
      "execution_count": 59,
      "outputs": []
    }
  ]
}