randomforest.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Malicious Domain Dataset - Random Forest Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    32613\n",
       "1     3815\n",
       "Name: Class, dtype: int64"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "# Load preprocessed dataset\n",
    "datasource = \"preprocessed.csv\"\n",
    "data = pd.read_csv(datasource)\n",
    "data.Class.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Resampling\n",
    "\n",
    "As the preprocessed dataset is unbalanced, we resample to create a balanced set for training and testing."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.utils import resample\n",
    "\n",
    "sample_size = min(data.Class.value_counts())\n",
    "\n",
    "malicious = resample(data.loc[data[\"Class\"] == 0], replace=False, n_samples=sample_size, random_state=42)\n",
    "benign = resample(data.loc[data[\"Class\"] == 1], replace=False, n_samples=sample_size, random_state=42)\n",
    "\n",
    "balanced = pd.concat([malicious, benign])\n",
    "balanced.Class.value_counts()\n",
    "\n",
    "X = balanced.drop(columns=[\"Class\"])\n",
    "y = balanced[\"Class\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Cross-validation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here we assess how well a Random Forest classifier will generalize the dataset by performing 10-fold cross-validation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "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>fit_time</th>\n",
       "      <th>score_time</th>\n",
       "      <th>test_accuracy</th>\n",
       "      <th>test_f1</th>\n",
       "      <th>test_precision</th>\n",
       "      <th>test_recall</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.239348</td>\n",
       "      <td>0.015891</td>\n",
       "      <td>0.967235</td>\n",
       "      <td>0.967062</td>\n",
       "      <td>0.970899</td>\n",
       "      <td>0.963255</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.238916</td>\n",
       "      <td>0.016546</td>\n",
       "      <td>0.973788</td>\n",
       "      <td>0.973958</td>\n",
       "      <td>0.966408</td>\n",
       "      <td>0.981627</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.227260</td>\n",
       "      <td>0.015876</td>\n",
       "      <td>0.973788</td>\n",
       "      <td>0.973684</td>\n",
       "      <td>0.976253</td>\n",
       "      <td>0.971129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.214280</td>\n",
       "      <td>0.015808</td>\n",
       "      <td>0.959371</td>\n",
       "      <td>0.960205</td>\n",
       "      <td>0.939698</td>\n",
       "      <td>0.981627</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.212008</td>\n",
       "      <td>0.015646</td>\n",
       "      <td>0.968545</td>\n",
       "      <td>0.968254</td>\n",
       "      <td>0.976000</td>\n",
       "      <td>0.960630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.209293</td>\n",
       "      <td>0.015539</td>\n",
       "      <td>0.969856</td>\n",
       "      <td>0.969777</td>\n",
       "      <td>0.973615</td>\n",
       "      <td>0.965969</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.210652</td>\n",
       "      <td>0.015744</td>\n",
       "      <td>0.965924</td>\n",
       "      <td>0.966408</td>\n",
       "      <td>0.954082</td>\n",
       "      <td>0.979058</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.211009</td>\n",
       "      <td>0.015825</td>\n",
       "      <td>0.969856</td>\n",
       "      <td>0.969856</td>\n",
       "      <td>0.971129</td>\n",
       "      <td>0.968586</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.215042</td>\n",
       "      <td>0.016558</td>\n",
       "      <td>0.971166</td>\n",
       "      <td>0.971204</td>\n",
       "      <td>0.971204</td>\n",
       "      <td>0.971204</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.213321</td>\n",
       "      <td>0.015398</td>\n",
       "      <td>0.982962</td>\n",
       "      <td>0.983007</td>\n",
       "      <td>0.981723</td>\n",
       "      <td>0.984293</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   fit_time  score_time  test_accuracy   test_f1  test_precision  test_recall\n",
       "0  0.239348    0.015891       0.967235  0.967062        0.970899     0.963255\n",
       "1  0.238916    0.016546       0.973788  0.973958        0.966408     0.981627\n",
       "2  0.227260    0.015876       0.973788  0.973684        0.976253     0.971129\n",
       "3  0.214280    0.015808       0.959371  0.960205        0.939698     0.981627\n",
       "4  0.212008    0.015646       0.968545  0.968254        0.976000     0.960630\n",
       "5  0.209293    0.015539       0.969856  0.969777        0.973615     0.965969\n",
       "6  0.210652    0.015744       0.965924  0.966408        0.954082     0.979058\n",
       "7  0.211009    0.015825       0.969856  0.969856        0.971129     0.968586\n",
       "8  0.215042    0.016558       0.971166  0.971204        0.971204     0.971204\n",
       "9  0.213321    0.015398       0.982962  0.983007        0.981723     0.984293"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.model_selection import cross_validate\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.feature_selection import SelectKBest\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "# Build pipeline to transform and fit data\n",
    "stages = [\n",
    "    (\"scaler\", StandardScaler()), # Standardize the data such that mean = 0 and variance = 1\n",
    "    (\"selector\", SelectKBest(k=6)), # Select the k best parameters using ANOVA F-value \n",
    "    (\"classifier\", RandomForestClassifier())] # Init Random Forest classifier\n",
    "\n",
    "pipeline = Pipeline(stages)\n",
    "\n",
    "# Perform cross validation using the pipeline\n",
    "folds = 10\n",
    "metrics = [\"accuracy\", \"f1\", \"precision\", \"recall\"]\n",
    "scores = cross_validate(pipeline, X, y, scoring=metrics, cv=folds)\n",
    "\n",
    "# Determine cross validation scores\n",
    "scores = pd.DataFrame.from_dict(scores)\n",
    "scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "fit_time          0.219113\n",
       "score_time        0.015883\n",
       "test_accuracy     0.970249\n",
       "test_f1           0.970342\n",
       "test_precision    0.968101\n",
       "test_recall       0.972738\n",
       "dtype: float64"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scores.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Finalized Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model accuraccy: 0.967453\n",
      "Selected features: ['StrangeCharacters' 'NumericRatio' 'VowelRatio' 'ConsonantSequence'\n",
      " 'NumericSequence' 'DomainLength']\n"
     ]
    }
   ],
   "source": [
    "# Partition data into testing and training sets\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, train_size=0.4, random_state=42)\n",
    "\n",
    "# Build pipeline to transform and fit data\n",
    "stages = [\n",
    "    (\"scaler\", StandardScaler()), # Standardize the data such that mean = 0 and variance = 1\n",
    "    (\"selector\", SelectKBest(k=6)), # Select the k best parameters using ANOVA F-value \n",
    "    (\"classifier\", RandomForestClassifier())] # Init Random Forest classifier\n",
    "\n",
    "pipeline = Pipeline(stages)\n",
    "\n",
    "pipeline.fit(X_train, y_train)\n",
    "\n",
    "# Score the resulting model and determine which features were selected\n",
    "score = pipeline.score(X_test, y_test)\n",
    "k_features = X.columns.values[pipeline['selector'].get_support()]\n",
    "\n",
    "print(\"Model accuraccy: %f\" % (score))\n",
    "print(\"Selected features: %s\" % (k_features))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualizations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.det_curve.DetCurveDisplay at 0x12baed360>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import ConfusionMatrixDisplay, RocCurveDisplay, DetCurveDisplay\n",
    "\n",
    "ConfusionMatrixDisplay.from_estimator(pipeline, X_test, y_test)\n",
    "\n",
    "RocCurveDisplay.from_estimator(pipeline, X_test, y_test)\n",
    "\n",
    "DetCurveDisplay.from_estimator(pipeline, X_test, y_test)"
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "aee8b7b246df8f9039afb4144a1f6fd8d2ca17a180786b69acc140d282b71a49"
  },
  "kernelspec": {
   "display_name": "Python 3.10.0 64-bit",
   "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": "3.10.0"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}