Phishing_Website_Detection_Models_and_Training.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "810e3386",
   "metadata": {},
   "source": [
    "# Phishing Website Detection by Machine Learning - Models & Training"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "358a7c81",
   "metadata": {},
   "source": [
    "This notebook aims to train machine learning models and deep neural nets on the dataset created to predict phishing websites. Both phishing and benign URLs of websites are gathered to form a dataset and from them required URL and website content-based features are extracted. The performance level of each model is measures and compared."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "82c575b6",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.metrics import accuracy_score\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from xgboost import XGBClassifier\n",
    "from sklearn.svm import SVC\n",
    "import pickle"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "469c88cb",
   "metadata": {},
   "source": [
    "# 1. Loading Data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "00ef9258",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Domain</th>\n",
       "      <th>Have_IP</th>\n",
       "      <th>Have_At</th>\n",
       "      <th>URL_Length</th>\n",
       "      <th>URL_Depth</th>\n",
       "      <th>Redirection</th>\n",
       "      <th>https_Domain</th>\n",
       "      <th>TinyURL</th>\n",
       "      <th>Prefix/Suffix</th>\n",
       "      <th>DNS_Record</th>\n",
       "      <th>Web_Traffic</th>\n",
       "      <th>Domain_Age</th>\n",
       "      <th>Domain_End</th>\n",
       "      <th>iFrame</th>\n",
       "      <th>Mouse_Over</th>\n",
       "      <th>Right_Click</th>\n",
       "      <th>Web_Forwards</th>\n",
       "      <th>Label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>graphicriver.net</td>\n",
       "      <td>False</td>\n",
       "      <td>False</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>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ecnavi.jp</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>hubpages.com</td>\n",
       "      <td>False</td>\n",
       "      <td>False</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>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>extratorrent.cc</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>3</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>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>icicibank.com</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>3</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>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Domain  Have_IP  Have_At  URL_Length  URL_Depth  Redirection  \\\n",
       "0  graphicriver.net    False    False           1          1            0   \n",
       "1         ecnavi.jp    False    False           1          1            1   \n",
       "2      hubpages.com    False    False           1          1            0   \n",
       "3   extratorrent.cc    False    False           1          3            0   \n",
       "4     icicibank.com    False    False           1          3            0   \n",
       "\n",
       "   https_Domain  TinyURL  Prefix/Suffix  DNS_Record  Web_Traffic  Domain_Age  \\\n",
       "0             0        0              0           1            1           1   \n",
       "1             0        0              0           1            1           1   \n",
       "2             0        0              0           1            1           1   \n",
       "3             0        0              0           1            0           1   \n",
       "4             0        0              0           1            1           1   \n",
       "\n",
       "   Domain_End  iFrame  Mouse_Over  Right_Click  Web_Forwards  Label  \n",
       "0           1       0           0            1             0      0  \n",
       "1           1       0           0            1             0      0  \n",
       "2           1       0           0            1             0      0  \n",
       "3           1       0           0            1             0      0  \n",
       "4           1       0           0            1             0      0  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_lp = pd.read_csv(\"Dataset//leg_phish-urls.csv\")\n",
    "df_lp.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "772b0603",
   "metadata": {},
   "source": [
    "# 2. Exploring Data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "662845e9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(10000, 18)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_lp.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "0a3bdc09",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Domain', 'Have_IP', 'Have_At', 'URL_Length', 'URL_Depth',\n",
       "       'Redirection', 'https_Domain', 'TinyURL', 'Prefix/Suffix', 'DNS_Record',\n",
       "       'Web_Traffic', 'Domain_Age', 'Domain_End', 'iFrame', 'Mouse_Over',\n",
       "       'Right_Click', 'Web_Forwards', 'Label'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_lp.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "5fb931d0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 10000 entries, 0 to 9999\n",
      "Data columns (total 18 columns):\n",
      " #   Column         Non-Null Count  Dtype \n",
      "---  ------         --------------  ----- \n",
      " 0   Domain         10000 non-null  object\n",
      " 1   Have_IP        10000 non-null  bool  \n",
      " 2   Have_At        10000 non-null  bool  \n",
      " 3   URL_Length     10000 non-null  int64 \n",
      " 4   URL_Depth      10000 non-null  int64 \n",
      " 5   Redirection    10000 non-null  int64 \n",
      " 6   https_Domain   10000 non-null  int64 \n",
      " 7   TinyURL        10000 non-null  int64 \n",
      " 8   Prefix/Suffix  10000 non-null  int64 \n",
      " 9   DNS_Record     10000 non-null  int64 \n",
      " 10  Web_Traffic    10000 non-null  int64 \n",
      " 11  Domain_Age     10000 non-null  int64 \n",
      " 12  Domain_End     10000 non-null  int64 \n",
      " 13  iFrame         10000 non-null  int64 \n",
      " 14  Mouse_Over     10000 non-null  int64 \n",
      " 15  Right_Click    10000 non-null  int64 \n",
      " 16  Web_Forwards   10000 non-null  int64 \n",
      " 17  Label          10000 non-null  int64 \n",
      "dtypes: bool(2), int64(15), object(1)\n",
      "memory usage: 1.2+ MB\n"
     ]
    }
   ],
   "source": [
    "df_lp.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "7a0880bf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Domain           0\n",
       "Have_IP          0\n",
       "Have_At          0\n",
       "URL_Length       0\n",
       "URL_Depth        0\n",
       "Redirection      0\n",
       "https_Domain     0\n",
       "TinyURL          0\n",
       "Prefix/Suffix    0\n",
       "DNS_Record       0\n",
       "Web_Traffic      0\n",
       "Domain_Age       0\n",
       "Domain_End       0\n",
       "iFrame           0\n",
       "Mouse_Over       0\n",
       "Right_Click      0\n",
       "Web_Forwards     0\n",
       "Label            0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_lp.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3492c0a4",
   "metadata": {},
   "source": [
    "# 3. Visualising Data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "1bdb427f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x1080 with 16 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_lp.hist(bins = 50, figsize = (15, 15))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "b8f1919a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x936 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (15, 13))\n",
    "sns.heatmap(df_lp.corr())\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f8277237",
   "metadata": {},
   "source": [
    "# 5. Data Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "3c9823ae",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>URL_Length</th>\n",
       "      <th>URL_Depth</th>\n",
       "      <th>Redirection</th>\n",
       "      <th>https_Domain</th>\n",
       "      <th>TinyURL</th>\n",
       "      <th>Prefix/Suffix</th>\n",
       "      <th>DNS_Record</th>\n",
       "      <th>Web_Traffic</th>\n",
       "      <th>Domain_Age</th>\n",
       "      <th>Domain_End</th>\n",
       "      <th>iFrame</th>\n",
       "      <th>Mouse_Over</th>\n",
       "      <th>Right_Click</th>\n",
       "      <th>Web_Forwards</th>\n",
       "      <th>Label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.00000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.0</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.0</td>\n",
       "      <td>10000.0</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.0000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.699800</td>\n",
       "      <td>2.558100</td>\n",
       "      <td>0.014200</td>\n",
       "      <td>0.00070</td>\n",
       "      <td>0.099300</td>\n",
       "      <td>0.113700</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.815200</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.090500</td>\n",
       "      <td>0.075000</td>\n",
       "      <td>0.9999</td>\n",
       "      <td>0.114500</td>\n",
       "      <td>0.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.458368</td>\n",
       "      <td>2.068732</td>\n",
       "      <td>0.118321</td>\n",
       "      <td>0.02645</td>\n",
       "      <td>0.299079</td>\n",
       "      <td>0.317462</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.388155</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.286911</td>\n",
       "      <td>0.263405</td>\n",
       "      <td>0.0100</td>\n",
       "      <td>0.318433</td>\n",
       "      <td>0.500025</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         URL_Length     URL_Depth   Redirection  https_Domain       TinyURL  \\\n",
       "count  10000.000000  10000.000000  10000.000000   10000.00000  10000.000000   \n",
       "mean       0.699800      2.558100      0.014200       0.00070      0.099300   \n",
       "std        0.458368      2.068732      0.118321       0.02645      0.299079   \n",
       "min        0.000000      0.000000      0.000000       0.00000      0.000000   \n",
       "25%        0.000000      1.000000      0.000000       0.00000      0.000000   \n",
       "50%        1.000000      2.000000      0.000000       0.00000      0.000000   \n",
       "75%        1.000000      4.000000      0.000000       0.00000      0.000000   \n",
       "max        1.000000     18.000000      1.000000       1.00000      1.000000   \n",
       "\n",
       "       Prefix/Suffix  DNS_Record   Web_Traffic  Domain_Age  Domain_End  \\\n",
       "count   10000.000000     10000.0  10000.000000     10000.0     10000.0   \n",
       "mean        0.113700         1.0      0.815200         1.0         1.0   \n",
       "std         0.317462         0.0      0.388155         0.0         0.0   \n",
       "min         0.000000         1.0      0.000000         1.0         1.0   \n",
       "25%         0.000000         1.0      1.000000         1.0         1.0   \n",
       "50%         0.000000         1.0      1.000000         1.0         1.0   \n",
       "75%         0.000000         1.0      1.000000         1.0         1.0   \n",
       "max         1.000000         1.0      1.000000         1.0         1.0   \n",
       "\n",
       "             iFrame    Mouse_Over  Right_Click  Web_Forwards         Label  \n",
       "count  10000.000000  10000.000000   10000.0000  10000.000000  10000.000000  \n",
       "mean       0.090500      0.075000       0.9999      0.114500      0.500000  \n",
       "std        0.286911      0.263405       0.0100      0.318433      0.500025  \n",
       "min        0.000000      0.000000       0.0000      0.000000      0.000000  \n",
       "25%        0.000000      0.000000       1.0000      0.000000      0.000000  \n",
       "50%        0.000000      0.000000       1.0000      0.000000      0.500000  \n",
       "75%        0.000000      0.000000       1.0000      0.000000      1.000000  \n",
       "max        1.000000      1.000000       1.0000      1.000000      1.000000  "
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_lp.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7e254f70",
   "metadata": {},
   "source": [
    "The Domain column doesnt have any significance to the machine learning model training. So dropping the 'Domain' column from the dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "c2ac7108",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Have_IP</th>\n",
       "      <th>Have_At</th>\n",
       "      <th>URL_Length</th>\n",
       "      <th>URL_Depth</th>\n",
       "      <th>Redirection</th>\n",
       "      <th>https_Domain</th>\n",
       "      <th>TinyURL</th>\n",
       "      <th>Prefix/Suffix</th>\n",
       "      <th>DNS_Record</th>\n",
       "      <th>Web_Traffic</th>\n",
       "      <th>Domain_Age</th>\n",
       "      <th>Domain_End</th>\n",
       "      <th>iFrame</th>\n",
       "      <th>Mouse_Over</th>\n",
       "      <th>Right_Click</th>\n",
       "      <th>Web_Forwards</th>\n",
       "      <th>Label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>False</td>\n",
       "      <td>False</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>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>False</td>\n",
       "      <td>False</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>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>3</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>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>3</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>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Have_IP  Have_At  URL_Length  URL_Depth  Redirection  https_Domain  \\\n",
       "0    False    False           1          1            0             0   \n",
       "1    False    False           1          1            1             0   \n",
       "2    False    False           1          1            0             0   \n",
       "3    False    False           1          3            0             0   \n",
       "4    False    False           1          3            0             0   \n",
       "\n",
       "   TinyURL  Prefix/Suffix  DNS_Record  Web_Traffic  Domain_Age  Domain_End  \\\n",
       "0        0              0           1            1           1           1   \n",
       "1        0              0           1            1           1           1   \n",
       "2        0              0           1            1           1           1   \n",
       "3        0              0           1            0           1           1   \n",
       "4        0              0           1            1           1           1   \n",
       "\n",
       "   iFrame  Mouse_Over  Right_Click  Web_Forwards  Label  \n",
       "0       0           0            1             0      0  \n",
       "1       0           0            1             0      0  \n",
       "2       0           0            1             0      0  \n",
       "3       0           0            1             0      0  \n",
       "4       0           0            1             0      0  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_lp = df_lp.drop(['Domain'], axis = 1).copy()\n",
    "df_lp.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "deda1c24",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_lp.loc[df_lp['Have_IP'] == True, 'Have_IP'] = 1\n",
    "df_lp.loc[df_lp['Have_IP'] == False, 'Have_IP'] = 0\n",
    "\n",
    "df_lp.loc[df_lp['Have_IP'] == True, 'Have_At'] = 1\n",
    "df_lp.loc[df_lp['Have_IP'] == False, 'Have_At'] = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "2641bb41",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Have_IP</th>\n",
       "      <th>Have_At</th>\n",
       "      <th>URL_Length</th>\n",
       "      <th>URL_Depth</th>\n",
       "      <th>Redirection</th>\n",
       "      <th>https_Domain</th>\n",
       "      <th>TinyURL</th>\n",
       "      <th>Prefix/Suffix</th>\n",
       "      <th>DNS_Record</th>\n",
       "      <th>Web_Traffic</th>\n",
       "      <th>Domain_Age</th>\n",
       "      <th>Domain_End</th>\n",
       "      <th>iFrame</th>\n",
       "      <th>Mouse_Over</th>\n",
       "      <th>Right_Click</th>\n",
       "      <th>Web_Forwards</th>\n",
       "      <th>Label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\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>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\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>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>3</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>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>3</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>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Have_IP Have_At  URL_Length  URL_Depth  Redirection  https_Domain  TinyURL  \\\n",
       "0       0       0           1          1            0             0        0   \n",
       "1       0       0           1          1            1             0        0   \n",
       "2       0       0           1          1            0             0        0   \n",
       "3       0       0           1          3            0             0        0   \n",
       "4       0       0           1          3            0             0        0   \n",
       "\n",
       "   Prefix/Suffix  DNS_Record  Web_Traffic  Domain_Age  Domain_End  iFrame  \\\n",
       "0              0           1            1           1           1       0   \n",
       "1              0           1            1           1           1       0   \n",
       "2              0           1            1           1           1       0   \n",
       "3              0           1            0           1           1       0   \n",
       "4              0           1            1           1           1       0   \n",
       "\n",
       "   Mouse_Over  Right_Click  Web_Forwards  Label  \n",
       "0           0            1             0      0  \n",
       "1           0            1             0      0  \n",
       "2           0            1             0      0  \n",
       "3           0            1             0      0  \n",
       "4           0            1             0      0  "
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_lp.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "943e7324",
   "metadata": {},
   "source": [
    "To even out the distribution while splitting the data into training & testing sets, we need to shuffle it. This even evades the case of overfitting while model training."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "024db55b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Have_IP</th>\n",
       "      <th>Have_At</th>\n",
       "      <th>URL_Length</th>\n",
       "      <th>URL_Depth</th>\n",
       "      <th>Redirection</th>\n",
       "      <th>https_Domain</th>\n",
       "      <th>TinyURL</th>\n",
       "      <th>Prefix/Suffix</th>\n",
       "      <th>DNS_Record</th>\n",
       "      <th>Web_Traffic</th>\n",
       "      <th>Domain_Age</th>\n",
       "      <th>Domain_End</th>\n",
       "      <th>iFrame</th>\n",
       "      <th>Mouse_Over</th>\n",
       "      <th>Right_Click</th>\n",
       "      <th>Web_Forwards</th>\n",
       "      <th>Label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</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>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\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>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</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>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>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>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Have_IP Have_At  URL_Length  URL_Depth  Redirection  https_Domain  TinyURL  \\\n",
       "0       0       0           1          2            0             0        0   \n",
       "1       0       0           1          3            1             0        0   \n",
       "2       0       0           1          1            0             0        0   \n",
       "3       0       0           0          2            0             0        0   \n",
       "4       0       0           0          0            0             0        0   \n",
       "\n",
       "   Prefix/Suffix  DNS_Record  Web_Traffic  Domain_Age  Domain_End  iFrame  \\\n",
       "0              0           1            1           1           1       1   \n",
       "1              0           1            1           1           1       0   \n",
       "2              0           1            0           1           1       0   \n",
       "3              0           1            1           1           1       0   \n",
       "4              0           1            1           1           1       0   \n",
       "\n",
       "   Mouse_Over  Right_Click  Web_Forwards  Label  \n",
       "0           1            1             1      0  \n",
       "1           0            1             0      1  \n",
       "2           0            1             0      1  \n",
       "3           0            1             0      1  \n",
       "4           0            1             0      1  "
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# shuffling the rows in the dataset so that when splitting the train and test set are equally distributed\n",
    "df_lp = df_lp.sample(frac=1).reset_index(drop=True)\n",
    "df_lp.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "545e9137",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 10000 entries, 0 to 9999\n",
      "Data columns (total 17 columns):\n",
      " #   Column         Non-Null Count  Dtype \n",
      "---  ------         --------------  ----- \n",
      " 0   Have_IP        10000 non-null  object\n",
      " 1   Have_At        10000 non-null  object\n",
      " 2   URL_Length     10000 non-null  int64 \n",
      " 3   URL_Depth      10000 non-null  int64 \n",
      " 4   Redirection    10000 non-null  int64 \n",
      " 5   https_Domain   10000 non-null  int64 \n",
      " 6   TinyURL        10000 non-null  int64 \n",
      " 7   Prefix/Suffix  10000 non-null  int64 \n",
      " 8   DNS_Record     10000 non-null  int64 \n",
      " 9   Web_Traffic    10000 non-null  int64 \n",
      " 10  Domain_Age     10000 non-null  int64 \n",
      " 11  Domain_End     10000 non-null  int64 \n",
      " 12  iFrame         10000 non-null  int64 \n",
      " 13  Mouse_Over     10000 non-null  int64 \n",
      " 14  Right_Click    10000 non-null  int64 \n",
      " 15  Web_Forwards   10000 non-null  int64 \n",
      " 16  Label          10000 non-null  int64 \n",
      "dtypes: int64(15), object(2)\n",
      "memory usage: 1.3+ MB\n"
     ]
    }
   ],
   "source": [
    "df_lp.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "902c3200",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_lp['Have_IP'] = pd.to_numeric(df_lp['Have_IP'])\n",
    "df_lp['Have_At'] = pd.to_numeric(df_lp['Have_At'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "72f64497",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 10000 entries, 0 to 9999\n",
      "Data columns (total 17 columns):\n",
      " #   Column         Non-Null Count  Dtype\n",
      "---  ------         --------------  -----\n",
      " 0   Have_IP        10000 non-null  int64\n",
      " 1   Have_At        10000 non-null  int64\n",
      " 2   URL_Length     10000 non-null  int64\n",
      " 3   URL_Depth      10000 non-null  int64\n",
      " 4   Redirection    10000 non-null  int64\n",
      " 5   https_Domain   10000 non-null  int64\n",
      " 6   TinyURL        10000 non-null  int64\n",
      " 7   Prefix/Suffix  10000 non-null  int64\n",
      " 8   DNS_Record     10000 non-null  int64\n",
      " 9   Web_Traffic    10000 non-null  int64\n",
      " 10  Domain_Age     10000 non-null  int64\n",
      " 11  Domain_End     10000 non-null  int64\n",
      " 12  iFrame         10000 non-null  int64\n",
      " 13  Mouse_Over     10000 non-null  int64\n",
      " 14  Right_Click    10000 non-null  int64\n",
      " 15  Web_Forwards   10000 non-null  int64\n",
      " 16  Label          10000 non-null  int64\n",
      "dtypes: int64(17)\n",
      "memory usage: 1.3 MB\n"
     ]
    }
   ],
   "source": [
    "df_lp.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "503e03aa",
   "metadata": {},
   "source": [
    "# 6. Splitting the Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "afdab899",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((10000, 16), (10000,))"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Separating & assigning features and target columns to X & y\n",
    "y = df_lp['Label']\n",
    "X = df_lp.drop('Label',axis=1)\n",
    "X.shape, y.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "0b5d0634",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((8000, 16), (2000, 16))"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Splitting the dataset into train and test sets: 80-20 split\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 12)\n",
    "X_train.shape, X_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "198f1d39",
   "metadata": {},
   "source": [
    "# 7. Machine Learning Models & Training"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "86381a5e",
   "metadata": {},
   "source": [
    "The supervised machine learning models (classification) considered to train the dataset in this notebook are:\n",
    "\n",
    "- Random Forest\n",
    "- XGBoost\n",
    "- Support Vector Machines"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "15501ab3",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Creating holders to store the model performance results\n",
    "ML_Model = []\n",
    "acc_train = []\n",
    "acc_test = []\n",
    "\n",
    "def storeResults(model, a,b):\n",
    "    ML_Model.append(model)\n",
    "    acc_train.append(round(a, 3))\n",
    "    acc_test.append(round(b, 3))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8e57e827",
   "metadata": {},
   "source": [
    "## 7.1. Random Forest:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b71d9f98",
   "metadata": {},
   "source": [
    "Random forests for regression and classification are currently among the most widely used machine learning methods.A random forest is essentially a collection of decision trees, where each tree is slightly different from the others. The idea behind random forests is that each tree might do a relatively good job of predicting, but will likely overfit on part of the data.\n",
    "\n",
    "If we build many trees, all of which work well and overfit in different ways, we can reduce the amount of overfitting by averaging their results. To build a random forest model, you need to decide on the number of trees to build (the n_estimators parameter of RandomForestRegressor or RandomForestClassifier). They are very powerful, often work well without heavy tuning of the parameters, and don’t require scaling of the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "97a5e1c0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(max_depth=5)"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# instantiate the model\n",
    "rForest = RandomForestClassifier(max_depth=5)\n",
    "\n",
    "# fit the model \n",
    "rForest.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "7b24fd49",
   "metadata": {},
   "outputs": [],
   "source": [
    "#predicting the target value from the model for the samples\n",
    "y_test_forest = rForest.predict(X_test)\n",
    "y_train_forest = rForest.predict(X_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5e82276e",
   "metadata": {},
   "source": [
    "### Performance Evaluation:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "1f1d474e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Random forest: Accuracy on training Data: 0.856\n",
      "Random forest: Accuracy on test Data: 0.849\n"
     ]
    }
   ],
   "source": [
    "#computing the accuracy of the model performance\n",
    "acc_train_forest = accuracy_score(y_train,y_train_forest)\n",
    "acc_test_forest = accuracy_score(y_test,y_test_forest)\n",
    "\n",
    "print(\"Random forest: Accuracy on training Data: {:.3f}\".format(acc_train_forest))\n",
    "print(\"Random forest: Accuracy on test Data: {:.3f}\".format(acc_test_forest))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "54d48221",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#checking the feature improtance in the model\n",
    "plt.figure(figsize=(9,7))\n",
    "n_features = X_train.shape[1]\n",
    "plt.barh(range(n_features), rForest.feature_importances_, align='center')\n",
    "plt.yticks(np.arange(n_features), X_train.columns)\n",
    "plt.xlabel(\"Feature importance\")\n",
    "plt.ylabel(\"Feature\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a08303b7",
   "metadata": {},
   "source": [
    "### Storing the results:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "33d54c88",
   "metadata": {},
   "outputs": [],
   "source": [
    "#storing the results. The below mentioned order of parameter passing is important.\n",
    "#Caution: Execute only once to avoid duplications.\n",
    "storeResults('Random Forest', acc_train_forest, acc_test_forest)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f97443cf",
   "metadata": {},
   "source": [
    "## 7.2. XGBoost:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "867b91dc",
   "metadata": {},
   "source": [
    "XGBoost is one of the most popular machine learning algorithms these days. XGBoost stands for eXtreme Gradient Boosting. Regardless of the type of prediction task at hand; regression or classification. XGBoost is an implementation of gradient boosted decision trees designed for speed and performance."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "0e4557d9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[01:25:28] WARNING: C:/Users/Administrator/workspace/xgboost-win64_release_1.5.0/src/learner.cc:1115: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'binary:logistic' was changed from 'error' to 'logloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "XGBClassifier(base_score=0.5, booster='gbtree', colsample_bylevel=1,\n",
       "              colsample_bynode=1, colsample_bytree=1, enable_categorical=False,\n",
       "              gamma=0, gpu_id=-1, importance_type=None,\n",
       "              interaction_constraints='', learning_rate=0.4, max_delta_step=0,\n",
       "              max_depth=7, min_child_weight=1, missing=nan,\n",
       "              monotone_constraints='()', n_estimators=100, n_jobs=4,\n",
       "              num_parallel_tree=1, predictor='auto', random_state=0,\n",
       "              reg_alpha=0, reg_lambda=1, scale_pos_weight=1, subsample=1,\n",
       "              tree_method='exact', use_label_encoder=False,\n",
       "              validate_parameters=1, verbosity=None)"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# instantiate the model\n",
    "xgb = XGBClassifier(use_label_encoder=False,learning_rate=0.4,max_depth=7)\n",
    "#fit the model\n",
    "xgb.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "d2767b5c",
   "metadata": {},
   "outputs": [],
   "source": [
    "#predicting the target value from the model for the samples\n",
    "y_test_xgb = xgb.predict(X_test)\n",
    "y_train_xgb = xgb.predict(X_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "357d445a",
   "metadata": {},
   "source": [
    "### Performance Evaluation:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "14487812",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "XGBoost: Accuracy on training Data: 0.871\n",
      "XGBoost : Accuracy on test Data: 0.867\n"
     ]
    }
   ],
   "source": [
    "#computing the accuracy of the model performance\n",
    "acc_train_xgb = accuracy_score(y_train,y_train_xgb)\n",
    "acc_test_xgb = accuracy_score(y_test,y_test_xgb)\n",
    "\n",
    "print(\"XGBoost: Accuracy on training Data: {:.3f}\".format(acc_train_xgb))\n",
    "print(\"XGBoost : Accuracy on test Data: {:.3f}\".format(acc_test_xgb))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "882f5e56",
   "metadata": {},
   "source": [
    "### Storing the results:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "976bd296",
   "metadata": {},
   "outputs": [],
   "source": [
    "#storing the results. The below mentioned order of parameter passing is important.\n",
    "#Caution: Execute only once to avoid duplications.\n",
    "storeResults('XGBoost', acc_train_xgb, acc_test_xgb)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1b34ff07",
   "metadata": {},
   "source": [
    "## 7.3. Support Vector Machines:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4e1e13de",
   "metadata": {},
   "source": [
    "In machine learning, support-vector machines (SVMs, also support-vector networks) are supervised learning models with associated learning algorithms that analyze data used for classification and regression analysis. Given a set of training examples, each marked as belonging to one or the other of two categories, an SVM training algorithm builds a model that assigns new examples to one category or the other, making it a non-probabilistic binary linear classifier."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "id": "626a7cc1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(kernel='linear', random_state=12)"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# instantiate the model\n",
    "svm = SVC(kernel='linear', C=1.0, random_state=12)\n",
    "#fit the model\n",
    "svm.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "id": "9f37dbcd",
   "metadata": {},
   "outputs": [],
   "source": [
    "#predicting the target value from the model for the samples\n",
    "y_test_svm = svm.predict(X_test)\n",
    "y_train_svm = svm.predict(X_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f18fc2d6",
   "metadata": {},
   "source": [
    "### Performance Evaluation:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "id": "51c1f9ae",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SVM: Accuracy on training Data: 0.852\n",
      "SVM : Accuracy on test Data: 0.849\n"
     ]
    }
   ],
   "source": [
    "#computing the accuracy of the model performance\n",
    "acc_train_svm = accuracy_score(y_train,y_train_svm)\n",
    "acc_test_svm = accuracy_score(y_test,y_test_svm)\n",
    "\n",
    "print(\"SVM: Accuracy on training Data: {:.3f}\".format(acc_train_svm))\n",
    "print(\"SVM : Accuracy on test Data: {:.3f}\".format(acc_test_svm))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "339e1f8c",
   "metadata": {},
   "source": [
    "### Storing the Results:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "8d36a71e",
   "metadata": {},
   "outputs": [],
   "source": [
    "#storing the results. The below mentioned order of parameter passing is important.\n",
    "#Caution: Execute only once to avoid duplications.\n",
    "storeResults('SVM', acc_train_svm, acc_test_svm)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "710ae963",
   "metadata": {},
   "source": [
    "# 8. Comparision of Models:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "928d6974",
   "metadata": {},
   "source": [
    "To compare the models performance, a dataframe is created. The columns of this dataframe are the lists created to store the results of the model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "id": "bca7fce5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ML Model</th>\n",
       "      <th>Train Accuracy</th>\n",
       "      <th>Test Accuracy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Random Forest</td>\n",
       "      <td>0.856</td>\n",
       "      <td>0.848</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.871</td>\n",
       "      <td>0.867</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>SVM</td>\n",
       "      <td>0.852</td>\n",
       "      <td>0.848</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        ML Model  Train Accuracy  Test Accuracy\n",
       "0  Random Forest           0.856          0.848\n",
       "1        XGBoost           0.871          0.867\n",
       "2            SVM           0.852          0.848"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#creating dataframe\n",
    "compResults = pd.DataFrame({ 'ML Model': ML_Model,    \n",
    "    'Train Accuracy': acc_train,\n",
    "    'Test Accuracy': acc_test})\n",
    "compResults"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "92958b5c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ML Model</th>\n",
       "      <th>Train Accuracy</th>\n",
       "      <th>Test Accuracy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.871</td>\n",
       "      <td>0.867</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Random Forest</td>\n",
       "      <td>0.856</td>\n",
       "      <td>0.848</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>SVM</td>\n",
       "      <td>0.852</td>\n",
       "      <td>0.848</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        ML Model  Train Accuracy  Test Accuracy\n",
       "1        XGBoost           0.871          0.867\n",
       "0  Random Forest           0.856          0.848\n",
       "2            SVM           0.852          0.848"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Sorting the datafram on accuracy\n",
    "compResults.sort_values(by=['Test Accuracy', 'Train Accuracy'], ascending=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "19bd3ea7",
   "metadata": {},
   "source": [
    "For the above comparision, it is clear that the XGBoost Classifier works well with this dataset.\n",
    "\n",
    "So, saving the model for future use."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "ab265560",
   "metadata": {},
   "outputs": [],
   "source": [
    "# save XGBoost model to file\n",
    "pickle.dump(xgb, open(\"XGBoostClassifier.pickle.dat\", \"wb\"))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "91a015e0",
   "metadata": {},
   "source": [
    "### Testing the saved model:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "id": "96b6164f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# load model from file\n",
    "loaded_model = pickle.load(open(\"XGBoostClassifier.pickle.dat\", \"rb\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "id": "f13e5281",
   "metadata": {},
   "outputs": [],
   "source": [
    "input = {\n",
    "#     0, 0, 1, 2, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1\n",
    "    \"Have_IP\" : 0,\n",
    "    \"Have_At\" : 0,\n",
    "    \"URL_Length\" : 1,  \n",
    "    \"URL_Depth\" : 2,\n",
    "    \"Redirection\" : 0,\n",
    "    \"https_Domain\" : 0,\n",
    "    \"TinyURL\" : 0,\n",
    "    \"Prefix/Suffix\" : 0,\n",
    "    \"DNS_Record\" : 1,\n",
    "    \"Web_Traffic\" : 1,\n",
    "    \"Domain_Age\": 1,\n",
    "    \"Domain_End\": 1,\n",
    "    \"iFrame\" : 1,\n",
    "    \"Mouse_Over\" : 1,\n",
    "    \"Right_Click\" : 1,\n",
    "    \"Web_Forwards\" : 1,\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "id": "fcf6789d",
   "metadata": {},
   "outputs": [],
   "source": [
    "input_df = pd.DataFrame(input,index=[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "id": "408519b5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.9758103 , 0.02418974]], dtype=float32)"
      ]
     },
     "execution_count": 102,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loaded_model.predict_proba(input_df)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ca4f6967",
   "metadata": {},
   "source": [
    "# 9. References:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7553a65f",
   "metadata": {},
   "source": [
    "- https://blog.keras.io/building-autoencoders-in-keras.html\n",
    "- https://en.wikipedia.org/wiki/Autoencoder\n",
    "- https://mc.ai/a-beginners-guide-to-build-stacked-autoencoder-and-tying-weights-with-it/\n",
    "- https://github.com/shreyagopal/t81_558_deep_learning/blob/master/t81_558_class_14_03_anomaly.ipynb\n",
    "- https://machinelearningmastery.com/save-gradient-boosting-models-xgboost-python/"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}