Intrusion_detection_DecisionTree_CSE-CIC-IDS2018.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import MinMaxScaler, StandardScaler\n",
    "from sklearn.metrics import classification_report\n",
    "from sklearn.svm import OneClassSVM\n",
    "from sklearn.pipeline import Pipeline\n",
    "import os\n",
    "import glob\n",
    "import csv\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn import preprocessing\n",
    "from sklearn import tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "os.chdir(\"D:/Data\")\n",
    "csv_data_path = 'D:/Data/'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(79999, 81)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "traindata = pd.read_csv('D:/Data/preprocessed_train_DT_2018.csv')\n",
    "testdata = pd.read_csv('D:/Data/preprocessed_test_DT_2018.csv')\n",
    "traindata.pop(\"Unnamed: 0\")\n",
    "testdata.pop(\"Unnamed: 0\")\n",
    "\n",
    "traindata.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(20000, 81)"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "testdata.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 79999 entries, 0 to 79998\n",
      "Data columns (total 80 columns):\n",
      " #   Column             Non-Null Count  Dtype  \n",
      "---  ------             --------------  -----  \n",
      " 0   Dst Port           79999 non-null  int64  \n",
      " 1   Flow Duration      79999 non-null  int64  \n",
      " 2   Tot Fwd Pkts       79999 non-null  int64  \n",
      " 3   Tot Bwd Pkts       79999 non-null  int64  \n",
      " 4   TotLen Fwd Pkts    79999 non-null  int64  \n",
      " 5   TotLen Bwd Pkts    79999 non-null  int64  \n",
      " 6   Fwd Pkt Len Max    79999 non-null  int64  \n",
      " 7   Fwd Pkt Len Min    79999 non-null  int64  \n",
      " 8   Fwd Pkt Len Mean   79999 non-null  float64\n",
      " 9   Fwd Pkt Len Std    79999 non-null  float64\n",
      " 10  Bwd Pkt Len Max    79999 non-null  int64  \n",
      " 11  Bwd Pkt Len Min    79999 non-null  int64  \n",
      " 12  Bwd Pkt Len Mean   79999 non-null  float64\n",
      " 13  Bwd Pkt Len Std    79999 non-null  float64\n",
      " 14  Flow Byts/s        79999 non-null  float64\n",
      " 15  Flow Pkts/s        79999 non-null  float64\n",
      " 16  Flow IAT Mean      79999 non-null  float64\n",
      " 17  Flow IAT Std       79999 non-null  float64\n",
      " 18  Flow IAT Max       79999 non-null  int64  \n",
      " 19  Flow IAT Min       79999 non-null  int64  \n",
      " 20  Fwd IAT Tot        79999 non-null  int64  \n",
      " 21  Fwd IAT Mean       79999 non-null  float64\n",
      " 22  Fwd IAT Std        79999 non-null  float64\n",
      " 23  Fwd IAT Max        79999 non-null  int64  \n",
      " 24  Fwd IAT Min        79999 non-null  int64  \n",
      " 25  Bwd IAT Tot        79999 non-null  int64  \n",
      " 26  Bwd IAT Mean       79999 non-null  float64\n",
      " 27  Bwd IAT Std        79999 non-null  float64\n",
      " 28  Bwd IAT Max        79999 non-null  int64  \n",
      " 29  Bwd IAT Min        79999 non-null  int64  \n",
      " 30  Fwd PSH Flags      79999 non-null  int64  \n",
      " 31  Bwd PSH Flags      79999 non-null  int64  \n",
      " 32  Fwd URG Flags      79999 non-null  int64  \n",
      " 33  Bwd URG Flags      79999 non-null  int64  \n",
      " 34  Fwd Header Len     79999 non-null  int64  \n",
      " 35  Bwd Header Len     79999 non-null  int64  \n",
      " 36  Fwd Pkts/s         79999 non-null  float64\n",
      " 37  Bwd Pkts/s         79999 non-null  float64\n",
      " 38  Pkt Len Min        79999 non-null  int64  \n",
      " 39  Pkt Len Max        79999 non-null  int64  \n",
      " 40  Pkt Len Mean       79999 non-null  float64\n",
      " 41  Pkt Len Std        79999 non-null  float64\n",
      " 42  Pkt Len Var        79999 non-null  float64\n",
      " 43  FIN Flag Cnt       79999 non-null  int64  \n",
      " 44  SYN Flag Cnt       79999 non-null  int64  \n",
      " 45  RST Flag Cnt       79999 non-null  int64  \n",
      " 46  PSH Flag Cnt       79999 non-null  int64  \n",
      " 47  ACK Flag Cnt       79999 non-null  int64  \n",
      " 48  URG Flag Cnt       79999 non-null  int64  \n",
      " 49  CWE Flag Count     79999 non-null  int64  \n",
      " 50  ECE Flag Cnt       79999 non-null  int64  \n",
      " 51  Down/Up Ratio      79999 non-null  int64  \n",
      " 52  Pkt Size Avg       79999 non-null  float64\n",
      " 53  Fwd Seg Size Avg   79999 non-null  float64\n",
      " 54  Bwd Seg Size Avg   79999 non-null  float64\n",
      " 55  Fwd Byts/b Avg     79999 non-null  int64  \n",
      " 56  Fwd Pkts/b Avg     79999 non-null  int64  \n",
      " 57  Fwd Blk Rate Avg   79999 non-null  int64  \n",
      " 58  Bwd Byts/b Avg     79999 non-null  int64  \n",
      " 59  Bwd Pkts/b Avg     79999 non-null  int64  \n",
      " 60  Bwd Blk Rate Avg   79999 non-null  int64  \n",
      " 61  Subflow Fwd Pkts   79999 non-null  int64  \n",
      " 62  Subflow Fwd Byts   79999 non-null  int64  \n",
      " 63  Subflow Bwd Pkts   79999 non-null  int64  \n",
      " 64  Subflow Bwd Byts   79999 non-null  int64  \n",
      " 65  Init Fwd Win Byts  79999 non-null  int64  \n",
      " 66  Init Bwd Win Byts  79999 non-null  int64  \n",
      " 67  Fwd Act Data Pkts  79999 non-null  int64  \n",
      " 68  Fwd Seg Size Min   79999 non-null  int64  \n",
      " 69  Active Mean        79999 non-null  float64\n",
      " 70  Active Std         79999 non-null  float64\n",
      " 71  Active Max         79999 non-null  int64  \n",
      " 72  Active Min         79999 non-null  int64  \n",
      " 73  Idle Mean          79999 non-null  float64\n",
      " 74  Idle Std           79999 non-null  float64\n",
      " 75  Idle Max           79999 non-null  int64  \n",
      " 76  Idle Min           79999 non-null  int64  \n",
      " 77  Protocol_0         79999 non-null  int64  \n",
      " 78  Protocol_6         79999 non-null  int64  \n",
      " 79  Protocol_17        79999 non-null  int64  \n",
      "dtypes: float64(24), int64(56)\n",
      "memory usage: 48.8 MB\n"
     ]
    }
   ],
   "source": [
    "Y = traindata.pop('Label')\n",
    "X = traindata.iloc[:,0:82]\n",
    "C = testdata.pop('Label')\n",
    "T = testdata.iloc[:,0:82]\n",
    "\n",
    "X.info(verbose=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 0, ..., 1, 1, 1], dtype=int64)"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "_traindata = np.array(X)\n",
    "_trainlabel = np.array(Y)\n",
    "_traindata = X.replace((np.inf, -np.inf, np.nan), 0).reset_index(drop=True)\n",
    "_testdata = np.array(T)\n",
    "_testlabel = np.array(C)\n",
    "_testdata = T.replace((np.inf, -np.inf, np.nan), 0).reset_index(drop=True)\n",
    "_trainlabel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Normalization des données d'apprentissage\n",
    "#données des features\n",
    "from sklearn import preprocessing\n",
    "_traindataNp = _traindata.values #returns a numpy array\n",
    "min_max_scaler = preprocessing.MinMaxScaler()\n",
    "_traindata_scaled = min_max_scaler.fit_transform(_traindataNp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "#données des tests\n",
    "_testdata_np = _testdata.values #returns a numpy array\n",
    "min_max_scaler = preprocessing.MinMaxScaler()\n",
    "_testdata_scaled = min_max_scaler.fit_transform(_testdata_np)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn import preprocessing\n",
    "from sklearn.pipeline import make_pipeline, Pipeline\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "from sklearn.pipeline import Pipeline\n",
    "# 5. Declare data preprocessing steps\n",
    "dec_tree = tree.DecisionTreeClassifier()\n",
    "pipe = Pipeline(steps=[('min_max_scaler', min_max_scaler),('dec_tree', dec_tree)])\n",
    "criterion = ['gini', 'entropy']\n",
    "max_depth = [2,4,6,8,10,12]\n",
    "# Add a dict of estimator and estimator related parameters in this list\n",
    "parameters = dict(\n",
    "                      dec_tree__criterion=criterion,\n",
    "                      dec_tree__max_depth=max_depth)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 10 folds for each of 12 candidates, totalling 120 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done  34 tasks      | elapsed:    9.3s\n",
      "[Parallel(n_jobs=-1)]: Done 120 out of 120 | elapsed:   43.2s finished\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=10,\n",
       "             estimator=Pipeline(steps=[('min_max_scaler', MinMaxScaler()),\n",
       "                                       ('dec_tree', DecisionTreeClassifier())]),\n",
       "             n_jobs=-1,\n",
       "             param_grid={'dec_tree__criterion': ['gini', 'entropy'],\n",
       "                         'dec_tree__max_depth': [2, 4, 6, 8, 10, 12]},\n",
       "             verbose=1)"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "clf = GridSearchCV(pipe, parameters, cv=10,verbose=1,n_jobs=-1)\n",
    "clf.fit(_traindata, _trainlabel)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True\n",
      "0.99755\n"
     ]
    }
   ],
   "source": [
    "print (clf.refit)\n",
    " \n",
    "# 9. Evaluate model pipeline on test data\n",
    "pred = clf.predict(_testdata)\n",
    "from sklearn.metrics import accuracy_score\n",
    "print(accuracy_score(_testlabel, pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['dt_intrusion_detection_binary_classification.pkl']"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#  Save model for future use\n",
    "from joblib import dump, load\n",
    "dump(clf, 'dt_intrusion_detection_binary_classification.pkl')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Criterion: entropy\n",
      "Best max_depth: 12\n",
      "\n",
      "DecisionTreeClassifier(criterion='entropy', max_depth=12)\n"
     ]
    }
   ],
   "source": [
    "print('Best Criterion:', clf.best_estimator_.get_params()['dec_tree__criterion'])\n",
    "print('Best max_depth:', clf.best_estimator_.get_params()['dec_tree__max_depth'])\n",
    "print(); print(clf.best_estimator_.get_params()['dec_tree'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      1.00      1.00      8617\n",
      "           1       1.00      1.00      1.00     11383\n",
      "\n",
      "    accuracy                           1.00     20000\n",
      "   macro avg       1.00      1.00      1.00     20000\n",
      "weighted avg       1.00      1.00      1.00     20000\n",
      "\n",
      "[[ 8609     8]\n",
      " [   41 11342]]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import confusion_matrix,classification_report\n",
    "\n",
    "cm = confusion_matrix(_testlabel, pred)\n",
    "print(classification_report(_testlabel, pred))\n",
    "print(cm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "clfDT = tree.DecisionTreeClassifier(class_weight=None, criterion='entropy', max_depth=12)\n",
    "clfDT = clfDT.fit(_traindata, _trainlabel)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9975\n"
     ]
    }
   ],
   "source": [
    "# 9. Evaluate model pipeline on test data\n",
    "pred = clfDT.predict(_testdata)\n",
    "\n",
    "\n",
    "from sklearn.metrics import accuracy_score\n",
    "print(accuracy_score(_testlabel, pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      1.00      1.00      8617\n",
      "           1       1.00      1.00      1.00     11383\n",
      "\n",
      "    accuracy                           1.00     20000\n",
      "   macro avg       1.00      1.00      1.00     20000\n",
      "weighted avg       1.00      1.00      1.00     20000\n",
      "\n",
      "[[ 8607    10]\n",
      " [   40 11343]]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import confusion_matrix,classification_report\n",
    "\n",
    "cm = confusion_matrix(_testlabel, pred)\n",
    "print(classification_report(_testlabel, pred))\n",
    "print(cm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Area under the ROC curve : 0.998964\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import roc_curve, auc\n",
    "\n",
    "probas = clfDT.predict_proba(_testdata)\n",
    "# probas est une matrice de deux colonnes avec la proabilités d'appartenance à chaque classe\n",
    "\n",
    "fpr, tpr, thresholds = roc_curve(_testlabel.ravel(), probas[:, 1])\n",
    "roc_auc = auc(fpr, tpr)\n",
    "print(\"Area under the ROC curve : %f\" % roc_auc)\n",
    "\n",
    "plt.plot(fpr, tpr, label='ROC curve (area = %0.2f)' % roc_auc)\n",
    "plt.plot([0, 1], [0, 1], 'k--')\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.0])\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('Receiver operating characteristic example')\n",
    "plt.legend(loc=\"lower right\");"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Méthode avec GA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting features with genetic algorithm.\n",
      "gen\tnevals\tavg                      \tstd                    \tmin                    \tmax                      \n",
      "0  \t50    \t[ 0.93579418 11.64      ]\t[0.08630556 5.42497926]\t[0.62987059 1.        ]\t[ 0.99792497 20.        ]\n",
      "1  \t38    \t[ 0.98768118 14.7       ]\t[0.01402055 2.83725219]\t[0.93852472 9.        ]\t[ 0.99794997 20.        ]\n",
      "2  \t32    \t[ 0.99648547 16.24      ]\t[0.00420147 2.11243935]\t[ 0.97908747 11.        ]\t[ 0.99794997 20.        ]\n",
      "3  \t29    \t[ 0.99789097 16.88      ]\t[5.94298279e-05 1.77358394e+00]\t[ 0.99767495 14.        ]\t[ 0.99796248 20.        ]\n",
      "4  \t21    \t[ 0.99789497 17.02      ]\t[2.12974021e-04 1.84921605e+00]\t[ 0.99642495 14.        ]\t[ 0.99796248 20.        ]\n",
      "5  \t27    \t[ 0.99626596 17.04      ]\t[0.00810512 1.6365818 ]        \t[ 0.95546223 15.        ]\t[ 0.99796248 20.        ]\n",
      "6  \t30    \t[ 0.99688897 17.26      ]\t[0.0058074  1.46710599]        \t[ 0.95768722 14.        ]\t[ 0.99798747 20.        ]\n",
      "7  \t30    \t[ 0.99791272 17.66      ]\t[4.90242043e-05 1.66865215e+00]\t[ 0.99778747 14.        ]\t[ 0.99798747 20.        ]\n",
      "8  \t28    \t[ 0.99703772 17.66      ]\t[0.00562129 1.79566144]        \t[ 0.95778723 14.        ]\t[ 0.99798747 20.        ]\n",
      "9  \t31    \t[ 0.99726722 18.38      ]\t[0.00291175 1.67200478]        \t[ 0.98016246 14.        ]\t[ 0.99798747 20.        ]\n",
      "10 \t25    \t[ 0.99750947 18.54      ]\t[0.00267244 1.60262285]        \t[ 0.97886246 14.        ]\t[ 0.99798747 20.        ]\n",
      "11 \t29    \t[ 0.99727547 18.56      ]\t[0.0029109  1.18591737]        \t[ 0.98017496 15.        ]\t[ 0.99798747 20.        ]\n",
      "12 \t31    \t[ 0.99792572 18.94      ]\t[5.40520521e-05 6.75573830e-01]\t[ 0.99779996 17.        ]\t[ 0.99798747 20.        ]\n",
      "13 \t33    \t[ 0.99779797 18.82      ]\t[0.00054179 0.38418745]        \t[ 0.99527494 18.        ]\t[ 0.99798747 19.        ]\n",
      "14 \t23    \t[ 0.99789797 18.9       ]\t[2.24244001e-04 3.60555128e-01]\t[ 0.99636245 17.        ]\t[ 0.99798747 19.        ]\n",
      "15 \t27    \t[ 0.99745672 18.78      ]\t[0.00268013 0.72910905]        \t[ 0.97899997 15.        ]\t[ 0.99798747 19.        ]\n",
      "16 \t27    \t[ 0.99789747 18.72      ]\t[2.29529075e-04 6.01331855e-01]\t[ 0.99633744 17.        ]\t[ 0.99798747 19.        ]\n",
      "17 \t25    \t[ 0.99751172 18.56      ]\t[0.00250032 0.85229103]        \t[ 0.98022496 15.        ]\t[ 0.99798747 19.        ]\n",
      "18 \t26    \t[ 0.99675397 18.6       ]\t[0.00823472 0.52915026]        \t[ 0.93911222 17.        ]\t[ 0.99798747 19.        ]\n",
      "19 \t27    \t[ 0.99661772 18.32      ]\t[0.00617322 0.73321211]        \t[ 0.95776223 16.        ]\t[ 0.99798747 19.        ]\n",
      "20 \t37    \t[ 0.99680146 18.14      ]\t[0.00595676 1.00019998]        \t[ 0.95652473 15.        ]\t[ 0.99798747 19.        ]\n",
      "21 \t25    \t[ 0.99791697 18.4       ]\t[4.73553919e-05 6.63324958e-01]\t[ 0.99778746 17.        ]\t[ 0.99798747 20.        ]\n",
      "22 \t24    \t[ 0.99793047 18.14      ]\t[7.01017156e-05 6.32771681e-01]\t[ 0.99762495 16.        ]\t[ 0.99798747 19.        ]\n",
      "23 \t33    \t[ 0.99769097 17.96      ]\t[0.00151419 0.77356318]        \t[ 0.98709993 15.        ]\t[ 0.99798747 19.        ]\n",
      "24 \t33    \t[ 0.99742422 17.9       ]\t[0.00280686 0.87749644]        \t[ 0.97807496 15.        ]\t[ 0.99798747 19.        ]\n",
      "25 \t33    \t[ 0.99784922 17.92      ]\t[4.02333864e-04 7.95989950e-01]\t[ 0.99546247 15.        ]\t[ 0.99798747 20.        ]\n",
      "GeneticSelectionCV(caching=True, crossover_independent_proba=0.5, cv=10,\n",
      "                   estimator=DecisionTreeClassifier(criterion='entropy',\n",
      "                                                    max_depth=12),\n",
      "                   max_features=20, n_gen_no_change=10, n_jobs=-1,\n",
      "                   n_population=50, scoring='accuracy', verbose=1)\n"
     ]
    }
   ],
   "source": [
    "from genetic_selection import GeneticSelectionCV\n",
    "from sklearn.model_selection import StratifiedKFold, GridSearchCV, RandomizedSearchCV\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn import datasets\n",
    "estimator = tree.DecisionTreeClassifier(class_weight=None, criterion='entropy', max_depth=12)\n",
    "cv = GeneticSelectionCV(estimator,\n",
    "                                  cv=10,\n",
    "                                  verbose=1,\n",
    "                                  scoring=\"accuracy\",\n",
    "                                  max_features=20,\n",
    "                                  n_population=50,\n",
    "                                  crossover_proba=0.5,\n",
    "                                  mutation_proba=0.2,\n",
    "                                  n_generations=40,\n",
    "                                  crossover_independent_proba=0.5,\n",
    "                                  mutation_independent_proba=0.05,\n",
    "                                  tournament_size=3,\n",
    "                                  n_gen_no_change=10,\n",
    "                                  caching=True,\n",
    "                                  n_jobs=-1)\n",
    "\n",
    "print(cv.fit(_traindata, _trainlabel))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Chosen Feats:   ['Dst Port', 'Tot Bwd Pkts', 'Fwd Pkt Len Std', 'Bwd Pkt Len Min', 'Bwd IAT Max', 'Bwd Pkts/s', 'Pkt Len Min', 'URG Flag Cnt', 'ECE Flag Cnt', 'Down/Up Ratio', 'Pkt Size Avg', 'Bwd Blk Rate Avg', 'Init Fwd Win Byts', 'Init Bwd Win Byts', 'Active Mean', 'Active Min', 'Protocol_6', 'Protocol_17']\n",
      "score 0.9979874749968747\n"
     ]
    }
   ],
   "source": [
    "genfeats = _traindata.columns[cv.support_]\n",
    "genfeats = list(genfeats)\n",
    "print(\"Chosen Feats:  \", genfeats)\n",
    "#Le dernier score MCC (Matthews correlation coefficient) comme score final du modèle, qui est le meilleur\n",
    "cv_score = cv.generation_scores_[-1]\n",
    "print(\"score\",cv_score) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Results\n",
      "======================\n",
      "Confusion Matrix:\n",
      "[[ 8613     4]\n",
      " [   37 11346]]\n",
      "Error Value:0.0020499999999999963\n",
      "Accuracy_Score:0.99795\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import confusion_matrix, zero_one_loss, accuracy_score,classification_report\n",
    "\n",
    "\n",
    "pred = cv.predict(_testdata)\n",
    "results = confusion_matrix(_testlabel, pred)\n",
    "error = zero_one_loss(_testlabel, pred)\n",
    "accuracy_score = accuracy_score(_testlabel, pred)\n",
    "print(\" Results\\n======================\\nConfusion Matrix:\\n{}\\nError Value:{}\\nAccuracy_Score:{}\\n\".format(results, error, accuracy_score))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "#apprentissage du Decision Tree avec la selection de variable de GA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0            9\n",
       "1           20\n",
       "2           22\n",
       "3           23\n",
       "4           25\n",
       "         ...  \n",
       "19995    99985\n",
       "19996    99987\n",
       "19997    99990\n",
       "19998    99993\n",
       "19999    99996\n",
       "Name: Unnamed: 0, Length: 20000, dtype: int64"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "traindata = pd.read_csv('D:/Data/preprocessed_train_DT_2018.csv')\n",
    "testdata = pd.read_csv('D:/Data/preprocessed_test_DT_2018.csv')\n",
    "traindata.pop(\"Unnamed: 0\")\n",
    "testdata.pop(\"Unnamed: 0\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "Y = traindata.pop('Label')\n",
    "X = traindata.iloc[:,0:82]\n",
    "X = traindata[['Dst Port', 'Tot Bwd Pkts', 'Fwd Pkt Len Std', 'Bwd Pkt Len Min', 'Bwd IAT Max', 'Bwd Pkts/s', 'Pkt Len Min', 'URG Flag Cnt', 'ECE Flag Cnt', 'Down/Up Ratio', 'Pkt Size Avg', 'Bwd Blk Rate Avg', 'Init Fwd Win Byts', 'Init Bwd Win Byts', 'Active Mean', 'Active Min', 'Protocol_6', 'Protocol_17']]\n",
    "C = testdata.pop('Label')\n",
    "T = testdata.iloc[:,0:82]\n",
    "T = testdata[['Dst Port', 'Tot Bwd Pkts', 'Fwd Pkt Len Std', 'Bwd Pkt Len Min', 'Bwd IAT Max', 'Bwd Pkts/s', 'Pkt Len Min', 'URG Flag Cnt', 'ECE Flag Cnt', 'Down/Up Ratio', 'Pkt Size Avg', 'Bwd Blk Rate Avg', 'Init Fwd Win Byts', 'Init Bwd Win Byts', 'Active Mean', 'Active Min', 'Protocol_6', 'Protocol_17']]\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "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>Dst Port</th>\n",
       "      <th>Tot Bwd Pkts</th>\n",
       "      <th>Fwd Pkt Len Std</th>\n",
       "      <th>Bwd Pkt Len Min</th>\n",
       "      <th>Bwd IAT Max</th>\n",
       "      <th>Bwd Pkts/s</th>\n",
       "      <th>Pkt Len Min</th>\n",
       "      <th>URG Flag Cnt</th>\n",
       "      <th>ECE Flag Cnt</th>\n",
       "      <th>Down/Up Ratio</th>\n",
       "      <th>Pkt Size Avg</th>\n",
       "      <th>Bwd Blk Rate Avg</th>\n",
       "      <th>Init Fwd Win Byts</th>\n",
       "      <th>Init Bwd Win Byts</th>\n",
       "      <th>Active Mean</th>\n",
       "      <th>Active Min</th>\n",
       "      <th>Protocol_6</th>\n",
       "      <th>Protocol_17</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>500</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>500</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>583.333333</td>\n",
       "      <td>0</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>4000407.0</td>\n",
       "      <td>4000407</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3389</td>\n",
       "      <td>8</td>\n",
       "      <td>196.550183</td>\n",
       "      <td>0</td>\n",
       "      <td>953120</td>\n",
       "      <td>1.957627</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>151.047619</td>\n",
       "      <td>0</td>\n",
       "      <td>8192</td>\n",
       "      <td>62644</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>443</td>\n",
       "      <td>0</td>\n",
       "      <td>17.897858</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>10.333333</td>\n",
       "      <td>0</td>\n",
       "      <td>257</td>\n",
       "      <td>-1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>139</td>\n",
       "      <td>3</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>6099854</td>\n",
       "      <td>0.326774</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>1024</td>\n",
       "      <td>8192</td>\n",
       "      <td>3080815.0</td>\n",
       "      <td>3080815</td>\n",
       "      <td>1</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>79994</th>\n",
       "      <td>80</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>2053</td>\n",
       "      <td>-1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>79995</th>\n",
       "      <td>80</td>\n",
       "      <td>4</td>\n",
       "      <td>283.533419</td>\n",
       "      <td>0</td>\n",
       "      <td>5000915</td>\n",
       "      <td>0.798844</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>516.222222</td>\n",
       "      <td>0</td>\n",
       "      <td>8192</td>\n",
       "      <td>220</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>79996</th>\n",
       "      <td>80</td>\n",
       "      <td>4</td>\n",
       "      <td>352.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>5005165</td>\n",
       "      <td>0.798316</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>342.125000</td>\n",
       "      <td>0</td>\n",
       "      <td>8192</td>\n",
       "      <td>222</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>79997</th>\n",
       "      <td>80</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>2053</td>\n",
       "      <td>-1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>79998</th>\n",
       "      <td>80</td>\n",
       "      <td>3</td>\n",
       "      <td>202.587759</td>\n",
       "      <td>0</td>\n",
       "      <td>10685</td>\n",
       "      <td>0.597880</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>282.375000</td>\n",
       "      <td>0</td>\n",
       "      <td>8192</td>\n",
       "      <td>219</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>79999 rows × 18 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       Dst Port  Tot Bwd Pkts  Fwd Pkt Len Std  Bwd Pkt Len Min  Bwd IAT Max  \\\n",
       "0           500             0         0.000000                0            0   \n",
       "1          3389             8       196.550183                0       953120   \n",
       "2             0             0         0.000000                0            0   \n",
       "3           443             0        17.897858                0            0   \n",
       "4           139             3         0.000000                0      6099854   \n",
       "...         ...           ...              ...              ...          ...   \n",
       "79994        80             0         0.000000                0            0   \n",
       "79995        80             4       283.533419                0      5000915   \n",
       "79996        80             4       352.000000                0      5005165   \n",
       "79997        80             0         0.000000                0            0   \n",
       "79998        80             3       202.587759                0        10685   \n",
       "\n",
       "       Bwd Pkts/s  Pkt Len Min  URG Flag Cnt  ECE Flag Cnt  Down/Up Ratio  \\\n",
       "0        0.000000          500             0             0              0   \n",
       "1        1.957627            0             0             1              0   \n",
       "2        0.000000            0             0             0              0   \n",
       "3        0.000000            0             0             0              0   \n",
       "4        0.326774            0             0             0              3   \n",
       "...           ...          ...           ...           ...            ...   \n",
       "79994    0.000000            0             0             0              0   \n",
       "79995    0.798844            0             0             1              0   \n",
       "79996    0.798316            0             0             1              1   \n",
       "79997    0.000000            0             0             0              0   \n",
       "79998    0.597880            0             0             1              0   \n",
       "\n",
       "       Pkt Size Avg  Bwd Blk Rate Avg  Init Fwd Win Byts  Init Bwd Win Byts  \\\n",
       "0        583.333333                 0                 -1                 -1   \n",
       "1        151.047619                 0               8192              62644   \n",
       "2          0.000000                 0                 -1                 -1   \n",
       "3         10.333333                 0                257                 -1   \n",
       "4          0.000000                 0               1024               8192   \n",
       "...             ...               ...                ...                ...   \n",
       "79994      0.000000                 0               2053                 -1   \n",
       "79995    516.222222                 0               8192                220   \n",
       "79996    342.125000                 0               8192                222   \n",
       "79997      0.000000                 0               2053                 -1   \n",
       "79998    282.375000                 0               8192                219   \n",
       "\n",
       "       Active Mean  Active Min  Protocol_6  Protocol_17  \n",
       "0        4000407.0     4000407           0            1  \n",
       "1              0.0           0           1            0  \n",
       "2              0.0           0           0            0  \n",
       "3              0.0           0           1            0  \n",
       "4        3080815.0     3080815           1            0  \n",
       "...            ...         ...         ...          ...  \n",
       "79994          0.0           0           1            0  \n",
       "79995          0.0           0           1            0  \n",
       "79996          0.0           0           1            0  \n",
       "79997          0.0           0           1            0  \n",
       "79998          0.0           0           1            0  \n",
       "\n",
       "[79999 rows x 18 columns]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "_traindata = np.array(X)\n",
    "_trainlabel = np.array(Y)\n",
    "_traindata = X.replace((np.inf, -np.inf, np.nan), 0).reset_index(drop=True)\n",
    "_testdata = np.array(T)\n",
    "_testlabel = np.array(C)\n",
    "_testdata = T.replace((np.inf, -np.inf, np.nan), 0).reset_index(drop=True)\n",
    "_traindata"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn import preprocessing\n",
    "from sklearn.pipeline import make_pipeline, Pipeline\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "from sklearn.pipeline import Pipeline\n",
    "# 5. Declare data preprocessing steps\n",
    "min_max_scaler = preprocessing.MinMaxScaler()\n",
    "dec_tree = tree.DecisionTreeClassifier()\n",
    "pipe = Pipeline(steps=[('min_max_scaler', min_max_scaler),('dec_tree', dec_tree)])\n",
    "criterion = ['gini', 'entropy']\n",
    "max_depth = [2,4,6,8,10,12]\n",
    "# Add a dict of estimator and estimator related parameters in this list\n",
    "parameters = dict(\n",
    "                      dec_tree__criterion=criterion,\n",
    "                      dec_tree__max_depth=max_depth)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 10 folds for each of 12 candidates, totalling 120 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done  34 tasks      | elapsed:    3.3s\n",
      "[Parallel(n_jobs=-1)]: Done 120 out of 120 | elapsed:    8.8s finished\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=10,\n",
       "             estimator=Pipeline(steps=[('min_max_scaler', MinMaxScaler()),\n",
       "                                       ('dec_tree', DecisionTreeClassifier())]),\n",
       "             n_jobs=-1,\n",
       "             param_grid={'dec_tree__criterion': ['gini', 'entropy'],\n",
       "                         'dec_tree__max_depth': [2, 4, 6, 8, 10, 12]},\n",
       "             verbose=1)"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf = GridSearchCV(pipe, parameters, cv=10,verbose=1,n_jobs=-1)\n",
    "clf.fit(_traindata, _trainlabel)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True\n",
      "0.998\n"
     ]
    }
   ],
   "source": [
    "print (clf.refit)\n",
    " \n",
    "# 9. Evaluate model pipeline on test data\n",
    "pred = clf.predict(_testdata)\n",
    "from sklearn.metrics import accuracy_score\n",
    "print(accuracy_score(_testlabel, pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Criterion: entropy\n",
      "Best max_depth: 10\n",
      "\n",
      "DecisionTreeClassifier(criterion='entropy', max_depth=10)\n"
     ]
    }
   ],
   "source": [
    "print('Best Criterion:', clf.best_estimator_.get_params()['dec_tree__criterion'])\n",
    "print('Best max_depth:', clf.best_estimator_.get_params()['dec_tree__max_depth'])\n",
    "print(); print(clf.best_estimator_.get_params()['dec_tree'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      1.00      1.00      8617\n",
      "           1       1.00      1.00      1.00     11383\n",
      "\n",
      "    accuracy                           1.00     20000\n",
      "   macro avg       1.00      1.00      1.00     20000\n",
      "weighted avg       1.00      1.00      1.00     20000\n",
      "\n",
      "[[ 8614     3]\n",
      " [   37 11346]]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import confusion_matrix,classification_report\n",
    "\n",
    "cm = confusion_matrix(_testlabel, pred)\n",
    "print(classification_report(_testlabel, pred))\n",
    "print(cm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}