Classification.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import sklearn\n",
    "from sklearn.model_selection import cross_validate\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn import preprocessing\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import roc_curve\n",
    "from sklearn.metrics import roc_auc_score\n",
    "from matplotlib import pyplot\n",
    "from sklearn.model_selection import validation_curve\n",
    "import numpy as np\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def mean_score(scoring):\n",
    "    return {i:j.mean() for i,j in scoring.items()}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Loading data.\n",
    "- Load data and shuffle it with specific seeding."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": "       1  2  3  4  5  6  7  8  9  10  ...  20  21  22  23  24  25  26  27  28  \\\n3558  -1  1 -1 -1 -1 -1  0  1  1   1  ...   1   1   1  -1   1   1   1   1   1   \n1783  -1  1 -1 -1 -1  1  0 -1  1  -1  ...  -1  -1  -1  -1   1   1   1  -1   1   \n14018 -1 -1  1 -1 -1 -1  0 -1  1   1  ...  -1  -1  -1  -1   1  -1   1  -1   1   \n3464  -1 -1 -1 -1 -1 -1  1  1  1   1  ...   1   1   1   1   1   1   1   1   1   \n9950  -1 -1 -1 -1 -1 -1 -1 -1  1   1  ...  -1  -1  -1   1   1  -1  -1  -1   1   \n\n       29  \n3558    1  \n1783    1  \n14018  -1  \n3464    1  \n9950   -1  \n\n[5 rows x 29 columns]",
      "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>1</th>\n      <th>2</th>\n      <th>3</th>\n      <th>4</th>\n      <th>5</th>\n      <th>6</th>\n      <th>7</th>\n      <th>8</th>\n      <th>9</th>\n      <th>10</th>\n      <th>...</th>\n      <th>20</th>\n      <th>21</th>\n      <th>22</th>\n      <th>23</th>\n      <th>24</th>\n      <th>25</th>\n      <th>26</th>\n      <th>27</th>\n      <th>28</th>\n      <th>29</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>3558</th>\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      <td>1</td>\n      <td>1</td>\n      <td>1</td>\n      <td>...</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>1</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>1783</th>\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      <td>-1</td>\n      <td>1</td>\n      <td>-1</td>\n      <td>...</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>1</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>14018</th>\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      <td>-1</td>\n      <td>1</td>\n      <td>1</td>\n      <td>...</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>1</td>\n      <td>-1</td>\n    </tr>\n    <tr>\n      <th>3464</th>\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>1</td>\n      <td>1</td>\n      <td>...</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>1</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>9950</th>\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>1</td>\n      <td>1</td>\n      <td>...</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>1</td>\n      <td>-1</td>\n    </tr>\n  </tbody>\n</table>\n<p>5 rows × 29 columns</p>\n</div>"
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv(\"fixed_values_ds.csv\")\n",
    "df.columns = ['1','2','3','4','5','6','7','8','9','10','11','12','13','14','15','16','17','18','19','20','21','22','23','24','25','26','27','28','29']\n",
    "df = sklearn.utils.shuffle(df)\n",
    "X = df.drop(\"29\",axis=1).values\n",
    "X = preprocessing.scale(X)\n",
    "y = df['29'].values\n",
    "df.head()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Evalution metrics \n",
    " - Specifying evaluation metrics for classification models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "scoring = {'accuracy': 'accuracy',\n",
    "           'recall': 'recall',\n",
    "           'precision': 'precision',\n",
    "           'f1': 'f1'}\n",
    "fold_count=10"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Classication models for phishing detection:\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'fit_time': 0.5088544607162475, 'score_time': 0.02754373550415039, 'test_accuracy': 0.9254902878657466, 'test_recall': 0.9279204948420375, 'test_precision': 0.9235822006191159, 'test_f1': 0.9256727941376587}\n"
     ]
    }
   ],
   "source": [
    "rforest_clf=RandomForestClassifier()\n",
    "cross_val_scores = cross_validate(rforest_clf, X, y, cv=fold_count, scoring=scoring)\n",
    "rforest_clf_score = mean_score(cross_val_scores)\n",
    "print(rforest_clf_score)\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "### ROC Curve\n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done 100 out of 100 | elapsed:    0.2s finished\n",
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done 100 out of 100 | elapsed:    0.0s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "No Skill: ROC AUC=0.500\n",
      "Logistic: ROC AUC=0.977\n"
     ]
    },
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "trainX, testX, trainy, testy = train_test_split(X, y, test_size=0.5, random_state=2)\n",
    "ns_probs = [0 for _ in range(len(testy))]\n",
    "model = RandomForestClassifier(min_samples_split=7, verbose=True)\n",
    "model.fit(trainX, trainy)\n",
    "lr_probs = model.predict_proba(testX)\n",
    "lr_probs = lr_probs[:, 1]\n",
    "ns_auc = roc_auc_score(testy, ns_probs)\n",
    "lr_auc = roc_auc_score(testy, lr_probs)\n",
    "print('No Skill: ROC AUC=%.3f' % (ns_auc))\n",
    "print('Logistic: ROC AUC=%.3f' % (lr_auc))\n",
    "ns_fpr, ns_tpr, _ = roc_curve(testy, ns_probs)\n",
    "lr_fpr, lr_tpr, _ = roc_curve(testy, lr_probs)\n",
    "pyplot.plot(ns_fpr, ns_tpr, linestyle='--', label='No Skill')\n",
    "pyplot.plot(lr_fpr, lr_tpr, marker='.', label='Logistic')\n",
    "pyplot.xlabel('False Positive Rate')\n",
    "pyplot.ylabel('True Positive Rate')\n",
    "pyplot.legend()\n",
    "pyplot.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "### Validation Curve\n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 504x504 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "param_range = np.arange(1, 250, 2)\n",
    "\n",
    "train_scores, test_scores = validation_curve(RandomForestClassifier(),\n",
    "                              X, y, param_name=\"n_estimators\", param_range=param_range,\n",
    "                              cv=4, scoring=\"accuracy\", n_jobs=-1)\n",
    "train_mean = np.mean(train_scores, axis=1)\n",
    "train_std = np.std(train_scores, axis=1)\n",
    "\n",
    "test_mean = np.mean(test_scores, axis=1)\n",
    "test_std = np.std(test_scores, axis=1)\n",
    "\n",
    "pyplot.subplots(1, figsize=(7, 7))\n",
    "pyplot.plot(param_range, train_mean, label=\"Training score\", color=\"black\")\n",
    "pyplot.plot(param_range, test_mean, label=\"Cross-validation score\", color=\"dimgrey\")\n",
    "\n",
    "pyplot.fill_between(param_range, train_mean - train_std, train_mean + train_std, color=\"gray\")\n",
    "pyplot.fill_between(param_range, test_mean - test_std, test_mean + test_std, color=\"gainsboro\")\n",
    "\n",
    "pyplot.title(\"Random Forest Based Validation Curve\")\n",
    "pyplot.xlabel(\"Number Of Trees\")\n",
    "pyplot.ylabel(\"Accuracy Score\")\n",
    "pyplot.tight_layout()\n",
    "pyplot.legend(loc=\"best\")\n",
    "pyplot.show()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "outputs": [],
   "source": [],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "outputs": [],
   "source": [],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  }
 ],
 "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": "3.9.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}