Modeling_Notebook.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import sqlalchemy\n",
    "\n",
    "import itertools\n",
    "import os\n",
    "from datetime import datetime\n",
    "\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "from sklearn.dummy import DummyClassifier\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.naive_bayes import MultinomialNB\n",
    "from sklearn.naive_bayes import ComplementNB\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "from xgboost import XGBClassifier\n",
    "\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.metrics import precision_score\n",
    "from sklearn.metrics import recall_score\n",
    "from sklearn.metrics import f1_score\n",
    "from sklearn.metrics import confusion_matrix\n",
    "\n",
    "\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "import seaborn as sns "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from functions import load_malicious_df, load_benign_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load & Combine Datasets"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Load malicious datasets from each protocol"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1279, 83)"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "malicious_HOPOPT = load_malicious_df(directory='HOPOPT', sample_size=1000)\n",
    "malicious_HOPOPT.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(9241, 83)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "malicious_TCP = load_malicious_df(directory='TCP', sample_size=1000)\n",
    "malicious_TCP.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(11000, 83)"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "malicious_UDP = load_malicious_df(directory='UDP', sample_size=1000)\n",
    "malicious_UDP.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Load benign datasets from each protocol for balanced dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1279, 83)"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bal_benign_HOPOPT = load_benign_df(directory='HOPOPT', sample_size=len(malicious_HOPOPT))\n",
    "bal_benign_HOPOPT.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(9241, 83)"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bal_benign_TCP = load_benign_df(directory='TCP', sample_size=len(malicious_TCP))\n",
    "bal_benign_TCP.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(11000, 83)"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bal_benign_UDP = load_benign_df(directory='UDP', sample_size=len(malicious_UDP))\n",
    "bal_benign_UDP.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Load benign datasets from each protocol for anomaly detection dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4310, 83)"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anom_benign_HOPOPT = load_benign_df(directory='HOPOPT', sample_size=200*len(malicious_HOPOPT))\n",
    "anom_benign_HOPOPT.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1848200, 83)"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anom_benign_TCP = load_benign_df(directory='TCP', sample_size=200*len(malicious_TCP))\n",
    "anom_benign_TCP.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(40295, 83)"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anom_benign_UDP = load_benign_df(directory='UDP', sample_size=200*len(malicious_UDP))\n",
    "anom_benign_UDP.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Combine into balanced and anomaly detection dataframes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(43040, 83)"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "balanced_df = pd.concat([malicious_HOPOPT, malicious_TCP, malicious_UDP, bal_benign_HOPOPT, bal_benign_TCP, bal_benign_UDP])\n",
    "balanced_df.reset_index(drop=True, inplace=True)\n",
    "balanced_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1914325, 83)"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anomaly_df = pd.concat([malicious_HOPOPT, malicious_TCP, malicious_UDP, anom_benign_HOPOPT, anom_benign_TCP, anom_benign_UDP])\n",
    "anomaly_df.reset_index(drop=True, inplace=True)\n",
    "anomaly_df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Change all benign lables to simply 'benign'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "bal_benign_filt = balanced_df['Malicious'] == 0\n",
    "balanced_df['Label'] = np.where(benign_filt, 'Benign', balanced_df['Label'])\n",
    "\n",
    "anom_benign_filt = anomaly_df['Malicious'] == 0\n",
    "anomaly_df['Label'] = np.where(anom_benign_filt, 'Benign', anomaly_df['Label'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Check target distributions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    21520\n",
       "0    21520\n",
       "Name: Malicious, dtype: int64"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "balanced_df['Malicious'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    1892805\n",
       "1      21520\n",
       "Name: Malicious, dtype: int64"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anomaly_df['Malicious'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Benign     21520\n",
       "TFTP        2291\n",
       "Syn         2210\n",
       "DNS         2195\n",
       "NTP         2156\n",
       "NetBIOS     2094\n",
       "SNMP        2091\n",
       "UDP         2064\n",
       "MSSQL       1687\n",
       "LDAP        1677\n",
       "Portmap     1555\n",
       "SSDP        1500\n",
       "Name: Label, dtype: int64"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "balanced_df['Label'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Benign     1892805\n",
       "TFTP          2291\n",
       "Syn           2210\n",
       "DNS           2195\n",
       "NTP           2156\n",
       "NetBIOS       2094\n",
       "SNMP          2091\n",
       "UDP           2064\n",
       "MSSQL         1687\n",
       "LDAP          1677\n",
       "Portmap       1555\n",
       "SSDP          1500\n",
       "Name: Label, dtype: int64"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anomaly_df['Label'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Load and combine files for from each protocol"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2558, 83)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bal_HOPOPT_df = load_balanced_df(directory='HOPOPT', sample_size=1000)\n",
    "bal_HOPOPT_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(18482, 83)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bal_TCP_df = load_balanced_df(directory='TCP', sample_size=1000)\n",
    "bal_TCP_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(22000, 83)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bal_UDP_df = load_balanced_df(directory='UDP', sample_size=1000)\n",
    "bal_UDP_df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Combine all protocols into balanced dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(43040, 83)"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "balanced_df = pd.concat([bal_HOPOPT_df, bal_TCP_df, bal_UDP_df])\n",
    "balanced_df.reset_index(drop=True, inplace=True)\n",
    "del bal_HOPOPT_df, bal_TCP_df, bal_UDP_df\n",
    "balanced_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Flow_ID', 'Flow_Duration', 'Fwd_Total_Pkts', 'Bwd_Total_Pkts',\n",
       "       'Fwd_Total_Bytes', 'Bwd_Total_Bytes', 'Fwd_Pkt_Length_Max',\n",
       "       'Fwd_Pkt_Length_Min', 'Fwd_Pkt_Length_Mean', 'Fwd_Pkt_Length_Std',\n",
       "       'Bwd_Pkt_Length_Max', 'Bwd_Pkt_Length_Min', 'Bwd_Pkt_Length_Mean',\n",
       "       'Bwd_Pkt_Length_Std', 'Flow_Bytes_Sec', 'Flow_Pkts_Sec',\n",
       "       'Flow_IAT_Mean', 'Flow_IAT_Std', 'Flow_IAT_Max', 'Flow_IAT_Min',\n",
       "       'Fwd_IAT_Total', 'Fwd_IAT_Mean', 'Fwd_IAT_Std', 'Fwd_IAT_Max',\n",
       "       'Fwd_IAT_Min', 'Bwd_IAT_Total', 'Bwd_IAT_Mean', 'Bwd_IAT_Std',\n",
       "       'Bwd_IAT_Max', 'Bwd_IAT_Min', 'Fwd_PSH_Flags', 'Bwd_PSH_Flags',\n",
       "       'Fwd_URG_Flags', 'Bwd_URG_Flags', 'Fwd_Header_Length',\n",
       "       'Bwd_Header_Length', 'Fwd_Pkts_Sec', 'Bwd_Pkts_Sec', 'Pkt_Length_Min',\n",
       "       'Pkt_Length_Max', 'Pkt_Length_Mean', 'Pkt_Length_Std', 'Pkt_Length_Var',\n",
       "       'FIN_Flag_Count', 'SYN_Flag_Count', 'RST_Flag_Count', 'PSH_Flag_Count',\n",
       "       'ACK_Flag_Count', 'URG_Flag_Count', 'CWE_Flag_Count', 'ECE_Flag_Count',\n",
       "       'Down_Up_Ratio', 'Pkt_Size_Mean', 'Fwd_Segment_Size_Mean',\n",
       "       'Bwd_Segment_Size_Mean', 'Fwd_Header_Length_1',\n",
       "       'Fwd_Byte_Bulk_Rate_Mean', 'Fwd_Pkt_Bulk_Rate_Mean',\n",
       "       'Fwd_Num_Bulk_Rate_Mean', 'Bwd_Byte_Bulk_Rate_Mean',\n",
       "       'Bwd_Pkt_Bulk_Rate_Mean', 'Bwd_Num_Bulk_Rate_Mean', 'Fwd_Subflow_Pkts',\n",
       "       'Fwd_Subflow_Bytes', 'Bwd_Subflow_Pkts', 'Bwd_Subflow_Bytes',\n",
       "       'Fwd_Init_Win_Bytes', 'Bwd_Init_Win_Bytes', 'Fwd_Act_Data_Pkt',\n",
       "       'Fwd_Seg_Size_Min', 'Time_Active_Mean', 'Time_Active_Std',\n",
       "       'Time_Active_Max', 'Time_Active_Min', 'Time_Idle_Mean', 'Time_Idle_Std',\n",
       "       'Time_Idle_Max', 'Time_Idle_Min', 'Label', 'HOPOPT', 'TCP', 'UDP',\n",
       "       'Malicious'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "balanced_df.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Change all benign lables to simply 'benign'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "benign_filt = balanced_df['Malicious'] == 0\n",
    "balanced_df['Label'] = np.where(benign_filt, 'Benign', balanced_df['Label'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Check target distributions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    21520\n",
       "0    21520\n",
       "Name: Malicious, dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "balanced_df['Malicious'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Benign     21520\n",
       "TFTP        2291\n",
       "Syn         2210\n",
       "DNS         2195\n",
       "NTP         2156\n",
       "NetBIOS     2094\n",
       "SNMP        2091\n",
       "UDP         2064\n",
       "MSSQL       1687\n",
       "LDAP        1677\n",
       "Portmap     1555\n",
       "SSDP        1500\n",
       "Name: Label, dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "balanced_df['Label'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Balanced Dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Binary Classification (Balanced)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train Test Split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [],
   "source": [
    "bal_binary_X = balanced_df.drop(['Flow_ID', 'Label', 'HOPOPT', 'Malicious'], axis=1)\n",
    "bal_binary_y = balanced_df['Malicious']\n",
    "\n",
    "bal_binary_X_train, bal_binary_X_test, bal_binary_y_train, bal_binary_y_test = train_test_split(\n",
    "    bal_binary_X, bal_binary_y, test_size = 0.25, random_state = 0)\n",
    "\n",
    "# Convert to list to avoid type error\n",
    "bal_binary_y_train = list(bal_binary_y_train)\n",
    "bal_binary_y_test = list(bal_binary_y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Standard Scalar"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [],
   "source": [
    "std_scaler = StandardScaler()\n",
    "\n",
    "bal_binary_X_train_scl = std_scaler.fit_transform(bal_binary_X_train)\n",
    "bal_binary_X_test_scl = std_scaler.transform(bal_binary_X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Dummy Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/eric/opt/anaconda3/lib/python3.7/site-packages/sklearn/dummy.py:132: FutureWarning: The default value of strategy will change from stratified to prior in 0.24.\n",
      "  \"stratified to prior in 0.24.\", FutureWarning)\n"
     ]
    }
   ],
   "source": [
    "bal_binary_dummy = DummyClassifier(random_state=0)\n",
    "bal_binary_dummy.fit(bal_binary_X_train_scl, bal_binary_y_train)\n",
    "bal_binary_dummy_preds = list(bal_binary_dummy.predict(bal_binary_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.5013940520446096\n",
      "      F1: 0.5049367906247116\n"
     ]
    }
   ],
   "source": [
    "bal_binary_dummy_acc = accuracy_score(bal_binary_y_test, bal_binary_dummy_preds)\n",
    "bal_binary_dummy_f1 = f1_score(bal_binary_y_test, bal_binary_dummy_preds)\n",
    "\n",
    "print('Accuracy:', bal_binary_dummy_acc)\n",
    "print('      F1:', bal_binary_dummy_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Gaussian Naive Bayes "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "bal_binary_gnb = GaussianNB()\n",
    "bal_binary_gnb.fit(bal_binary_X_train_scl, bal_binary_y_train)\n",
    "bal_binary_gnb_preds = list(bal_binary_gnb.predict(bal_binary_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.6943308550185874\n",
      "      F1: 0.749600304529882\n"
     ]
    }
   ],
   "source": [
    "bal_binary_gnb_acc = accuracy_score(bal_binary_y_test, bal_binary_gnb_preds)\n",
    "bal_binary_gnb_f1 = f1_score(bal_binary_y_test, bal_binary_gnb_preds)\n",
    "\n",
    "print('Accuracy:', bal_binary_gnb_acc)\n",
    "print('      F1:', bal_binary_gnb_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Decision Tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "bal_binary_dt = DecisionTreeClassifier(random_state=0)\n",
    "bal_binary_dt.fit(bal_binary_X_train_scl, bal_binary_y_train)\n",
    "bal_binary_dt_preds = list(bal_binary_dt.predict(bal_binary_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.9979553903345725\n",
      "      F1: 0.9979749631811488\n"
     ]
    }
   ],
   "source": [
    "bal_binary_dt_acc = accuracy_score(bal_binary_y_test, bal_binary_dt_preds)\n",
    "bal_binary_dt_f1 = f1_score(bal_binary_y_test, bal_binary_dt_preds)\n",
    "\n",
    "print('Accuracy:', bal_binary_dt_acc)\n",
    "print('      F1:', bal_binary_dt_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### K Nearest Neighbors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "bal_binary_knn = KNeighborsClassifier()\n",
    "bal_binary_knn.fit(bal_binary_X_train_scl, bal_binary_y_train)\n",
    "bal_binary_knn_preds = list(bal_binary_knn.predict(bal_binary_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.9960037174721189\n",
      "      F1: 0.9960416091319156\n"
     ]
    }
   ],
   "source": [
    "bal_binary_knn_acc = accuracy_score(bal_binary_y_test, bal_binary_knn_preds)\n",
    "bal_binary_knn_f1 = f1_score(bal_binary_y_test, bal_binary_knn_preds)\n",
    "\n",
    "print('Accuracy:', bal_binary_knn_acc)\n",
    "print('      F1:', bal_binary_knn_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "bal_binary_rf = RandomForestClassifier(random_state=0, n_jobs=-1)\n",
    "bal_binary_rf.fit(bal_binary_X_train_scl, bal_binary_y_train)\n",
    "bal_binary_rf_preds = list(bal_binary_rf.predict(bal_binary_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.9988847583643122\n",
      "      F1: 0.9988954344624447\n"
     ]
    }
   ],
   "source": [
    "bal_binary_rf_acc = accuracy_score(bal_binary_y_test, bal_binary_rf_preds)\n",
    "bal_binary_rf_f1 = f1_score(bal_binary_y_test, bal_binary_rf_preds)\n",
    "\n",
    "print('Accuracy:', bal_binary_rf_acc)\n",
    "print('      F1:', bal_binary_rf_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### XGBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "bal_binary_xgb = XGBClassifier(random_state=0, n_jobs=-1)\n",
    "bal_binary_xgb.fit(bal_binary_X_train_scl, bal_binary_y_train)\n",
    "bal_binary_xgb_preds = list(bal_binary_xgb.predict(bal_binary_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.9994423791821562\n",
      "      F1: 0.9994477172312224\n"
     ]
    }
   ],
   "source": [
    "bal_binary_xgb_acc = accuracy_score(bal_binary_y_test, bal_binary_xgb_preds)\n",
    "bal_binary_xgb_f1 = f1_score(bal_binary_y_test, bal_binary_xgb_preds)\n",
    "\n",
    "print('Accuracy:', bal_binary_xgb_acc)\n",
    "print('      F1:', bal_binary_xgb_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature Importances"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Decision Tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('Bwd_Pkt_Length_Max', 0.4321718724387804),\n",
       " ('Fwd_IAT_Min', 0.09638106984906894),\n",
       " ('ACK_Flag_Count', 0.09033950805271436),\n",
       " ('Bwd_Pkts_Sec', 0.04781227881326841),\n",
       " ('Pkt_Size_Mean', 0.047542461074881436),\n",
       " ('Fwd_Pkt_Length_Std', 0.047491182791384416),\n",
       " ('Fwd_Init_Win_Bytes', 0.041885485142335525),\n",
       " ('Bwd_Header_Length', 0.03417186684692964),\n",
       " ('Bwd_Init_Win_Bytes', 0.029978998366472578),\n",
       " ('URG_Flag_Count', 0.028887571059214186)]"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bal_binary_dt_features = dict(zip(bal_binary_X.columns, bal_binary_dt.feature_importances_))\n",
    "bal_binary_dt_features = sorted(bal_binary_dt_features.items(), key=lambda x: x[1], reverse=True)\n",
    "bal_binary_dt_features[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('Bwd_Segment_Size_Mean', 0.06479943731326662),\n",
       " ('Bwd_Pkt_Length_Max', 0.05768429926659383),\n",
       " ('Fwd_Init_Win_Bytes', 0.051959549429500464),\n",
       " ('Bwd_Pkt_Length_Mean', 0.051723724983398243),\n",
       " ('Pkt_Length_Var', 0.05011734358632702),\n",
       " ('Pkt_Length_Min', 0.03474782546490374),\n",
       " ('Bwd_Subflow_Bytes', 0.03389031994987526),\n",
       " ('Fwd_Pkt_Length_Min', 0.031065625410276246),\n",
       " ('Bwd_Total_Bytes', 0.02808381988258205),\n",
       " ('Pkt_Length_Std', 0.026423405011124897)]"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bal_binary_rf_features = dict(zip(bal_binary_X.columns, bal_binary_rf.feature_importances_))\n",
    "bal_binary_rf_features = sorted(bal_binary_rf_features.items(), key=lambda x: x[1], reverse=True)\n",
    "bal_binary_rf_features[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "XGBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('Bwd_Pkt_Length_Max', 0.7066348),\n",
       " ('Bwd_Pkt_Length_Mean', 0.046133853),\n",
       " ('Fwd_Pkt_Length_Std', 0.02572131),\n",
       " ('ACK_Flag_Count', 0.023738889),\n",
       " ('Bwd_Pkts_Sec', 0.018046636),\n",
       " ('Bwd_Header_Length', 0.01647789),\n",
       " ('PSH_Flag_Count', 0.015690256),\n",
       " ('Fwd_IAT_Min', 0.015220775),\n",
       " ('Fwd_Total_Pkts', 0.015060951),\n",
       " ('Pkt_Size_Mean', 0.011617822)]"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bal_binary_xgb_features = dict(zip(bal_binary_X.columns, bal_binary_xgb.feature_importances_))\n",
    "bal_binary_xgb_features = sorted(bal_binary_xgb_features.items(), key=lambda x: x[1], reverse=True)\n",
    "bal_binary_xgb_features[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Multiclass Classification (Balanced)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train Test Split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "bal_multi_X = balanced_df.drop(['Flow_ID', 'Label', 'HOPOPT', 'Malicious'], axis=1)\n",
    "bal_multi_y = balanced_df['Label']\n",
    "\n",
    "bal_multi_X_train, bal_multi_X_test, bal_multi_y_train, bal_multi_y_test = train_test_split(\n",
    "    bal_multi_X, bal_multi_y, test_size = 0.25, random_state = 0)\n",
    "\n",
    "# Convert to list to avoid type error\n",
    "bal_multi_y_train = list(bal_multi_y_train)\n",
    "bal_multi_y_test = list(bal_multi_y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Standard Scalar"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "std_scaler = StandardScaler()\n",
    "\n",
    "bal_multi_X_train_scl = std_scaler.fit_transform(bal_multi_X_train)\n",
    "bal_multi_X_test_scl = std_scaler.transform(bal_multi_X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Dummy Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/eric/opt/anaconda3/lib/python3.7/site-packages/sklearn/dummy.py:132: FutureWarning: The default value of strategy will change from stratified to prior in 0.24.\n",
      "  \"stratified to prior in 0.24.\", FutureWarning)\n"
     ]
    }
   ],
   "source": [
    "bal_multi_dummy = DummyClassifier(random_state=0)\n",
    "bal_multi_dummy.fit(bal_multi_X_train_scl, bal_multi_y_train)\n",
    "bal_multi_dummy_preds = list(bal_multi_dummy.predict(bal_multi_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.2776951672862453\n",
      "      F1: 0.27377289188976395\n"
     ]
    }
   ],
   "source": [
    "bal_multi_dummy_acc = accuracy_score(bal_multi_y_test, bal_multi_dummy_preds)\n",
    "bal_multi_dummy_f1 = f1_score(bal_multi_y_test, bal_multi_dummy_preds, average='weighted')\n",
    "\n",
    "print('Accuracy:', bal_multi_dummy_acc)\n",
    "print('      F1:', bal_multi_dummy_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Gaussian Naive Bayes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "bal_multi_gnb = GaussianNB()\n",
    "bal_multi_gnb.fit(bal_multi_X_train_scl, bal_multi_y_train)\n",
    "bal_multi_gnb_preds = list(bal_multi_gnb.predict(bal_multi_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.3140334572490706\n",
      "      F1: 0.3481777387956166\n"
     ]
    }
   ],
   "source": [
    "bal_multi_gnb_acc = accuracy_score(bal_multi_y_test, bal_multi_gnb_preds)\n",
    "bal_multi_gnb_f1 = f1_score(bal_multi_y_test, bal_multi_gnb_preds, average='weighted')\n",
    "\n",
    "print('Accuracy:', bal_multi_gnb_acc)\n",
    "print('      F1:', bal_multi_gnb_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Decision Tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "bal_multi_dt = DecisionTreeClassifier(random_state=0)\n",
    "bal_multi_dt.fit(bal_multi_X_train_scl, bal_multi_y_train)\n",
    "bal_multi_dt_preds = list(bal_multi_dt.predict(bal_multi_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.8467472118959107\n",
      "      F1: 0.8472017627615264\n"
     ]
    }
   ],
   "source": [
    "bal_multi_dt_acc = accuracy_score(bal_multi_y_test, bal_multi_dt_preds)\n",
    "bal_multi_dt_f1 = f1_score(bal_multi_y_test, bal_multi_dt_preds, average='weighted')\n",
    "\n",
    "print('Accuracy:', bal_multi_dt_acc)\n",
    "print('      F1:', bal_multi_dt_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### K Nearest Neighbors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "bal_multi_knn = KNeighborsClassifier()\n",
    "bal_multi_knn.fit(bal_multi_X_train_scl, bal_multi_y_train)\n",
    "bal_multi_knn_preds = list(bal_multi_knn.predict(bal_multi_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.8433085501858736\n",
      "      F1: 0.8396681128702419\n"
     ]
    }
   ],
   "source": [
    "bal_multi_knn_acc = accuracy_score(bal_multi_y_test, bal_multi_knn_preds)\n",
    "bal_multi_knn_f1 = f1_score(bal_multi_y_test, bal_multi_knn_preds, average='weighted')\n",
    "\n",
    "print('Accuracy:', bal_multi_knn_acc)\n",
    "print('      F1:', bal_multi_knn_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "bal_multi_rf = RandomForestClassifier(random_state=0, n_jobs=-1)\n",
    "bal_multi_rf.fit(bal_multi_X_train_scl, bal_multi_y_train)\n",
    "bal_multi_rf_preds = list(bal_multi_rf.predict(bal_multi_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.8523234200743495\n",
      "      F1: 0.8519008118496785\n"
     ]
    }
   ],
   "source": [
    "bal_multi_rf_acc = accuracy_score(bal_multi_y_test, bal_multi_rf_preds)\n",
    "bal_multi_rf_f1 = f1_score(bal_multi_y_test, bal_multi_rf_preds, average='weighted')\n",
    "\n",
    "print('Accuracy:', bal_multi_rf_acc)\n",
    "print('      F1:', bal_multi_rf_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### XGBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "bal_multi_xgb = XGBClassifier(random_state=0, n_jobs=-1)\n",
    "bal_multi_xgb.fit(bal_multi_X_train_scl, bal_multi_y_train)\n",
    "bal_multi_xgb_preds = list(bal_multi_xgb.predict(bal_multi_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.8650557620817844\n",
      "      F1: 0.8647730500675536\n"
     ]
    }
   ],
   "source": [
    "bal_multi_xgb_acc = accuracy_score(bal_multi_y_test, bal_multi_xgb_preds)\n",
    "bal_multi_xgb_f1 = f1_score(bal_multi_y_test, bal_multi_xgb_preds, average='weighted')\n",
    "\n",
    "print('Accuracy:', bal_multi_xgb_acc)\n",
    "print('      F1:', bal_multi_xgb_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature Importances"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Decision Tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('Bwd_Pkt_Length_Max', 0.17813608910486856),\n",
       " ('Pkt_Length_Min', 0.13232021964005997),\n",
       " ('Fwd_Init_Win_Bytes', 0.07129227402814149),\n",
       " ('Flow_Bytes_Sec', 0.05254508982369269),\n",
       " ('ACK_Flag_Count', 0.04822122872673132),\n",
       " ('Fwd_IAT_Min', 0.048125155685704676),\n",
       " ('Flow_IAT_Max', 0.04057134077946873),\n",
       " ('Bwd_Pkts_Sec', 0.02851069734541092),\n",
       " ('Fwd_Subflow_Bytes', 0.027439768028378247),\n",
       " ('Bwd_Header_Length', 0.027045377146699245)]"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bal_multi_dt_features = dict(zip(bal_multi_X.columns, bal_multi_dt.feature_importances_))\n",
    "bal_multi_dt_features = sorted(bal_multi_dt_features.items(), key=lambda x: x[1], reverse=True)\n",
    "bal_multi_dt_features[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('Fwd_Init_Win_Bytes', 0.052314072135577656),\n",
       " ('Pkt_Size_Mean', 0.033927197406239375),\n",
       " ('Pkt_Length_Min', 0.03362226368219828),\n",
       " ('Flow_IAT_Mean', 0.03229548261837191),\n",
       " ('Fwd_Pkt_Length_Min', 0.02929883891176453),\n",
       " ('Fwd_Segment_Size_Mean', 0.029189954369181214),\n",
       " ('Pkt_Length_Mean', 0.02841634731314821),\n",
       " ('Bwd_Segment_Size_Mean', 0.027442983138788266),\n",
       " ('Fwd_Pkts_Sec', 0.02733519211706754),\n",
       " ('Flow_IAT_Max', 0.02732665349636803)]"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bal_multi_rf_features = dict(zip(bal_multi_X.columns, bal_multi_rf.feature_importances_))\n",
    "bal_multi_rf_features = sorted(bal_multi_rf_features.items(), key=lambda x: x[1], reverse=True)\n",
    "bal_multi_rf_features[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "XGBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('Bwd_Pkt_Length_Max', 0.8722749),\n",
       " ('Pkt_Length_Min', 0.014750869),\n",
       " ('ACK_Flag_Count', 0.01015347),\n",
       " ('CWE_Flag_Count', 0.008745772),\n",
       " ('Time_Idle_Std', 0.0071296976),\n",
       " ('PSH_Flag_Count', 0.005871469),\n",
       " ('Bwd_Header_Length', 0.0050045243),\n",
       " ('Fwd_Total_Bytes', 0.0049487054),\n",
       " ('UDP', 0.004383982),\n",
       " ('URG_Flag_Count', 0.0042051356)]"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bal_multi_xgb_features = dict(zip(bal_multi_X.columns, bal_multi_xgb.feature_importances_))\n",
    "bal_multi_xgb_features = sorted(bal_multi_xgb_features.items(), key=lambda x: x[1], reverse=True)\n",
    "bal_multi_xgb_features[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Confusion Matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Normalized confusion matrix\n",
      "[[9.99624765e-01 0.00000000e+00 1.87617261e-04 0.00000000e+00\n",
      "  0.00000000e+00 0.00000000e+00 1.87617261e-04 0.00000000e+00\n",
      "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
      " [1.70648464e-03 7.06484642e-01 2.21843003e-02 3.92491468e-02\n",
      "  6.14334471e-02 2.73037543e-02 8.53242321e-03 4.60750853e-02\n",
      "  4.09556314e-02 0.00000000e+00 1.70648464e-03 4.43686007e-02]\n",
      " [0.00000000e+00 6.50000000e-02 5.77500000e-01 2.50000000e-01\n",
      "  1.25000000e-02 2.00000000e-02 1.50000000e-02 3.25000000e-02\n",
      "  1.50000000e-02 5.00000000e-03 0.00000000e+00 7.50000000e-03]\n",
      " [0.00000000e+00 7.21649485e-02 3.32474227e-01 3.73711340e-01\n",
      "  6.44329897e-02 3.35051546e-02 2.83505155e-02 2.31958763e-02\n",
      "  3.35051546e-02 5.15463918e-03 0.00000000e+00 3.35051546e-02]\n",
      " [1.82481752e-03 4.19708029e-02 3.64963504e-03 7.29927007e-03\n",
      "  7.73722628e-01 5.47445255e-03 5.47445255e-03 4.74452555e-02\n",
      "  7.29927007e-03 0.00000000e+00 3.64963504e-03 1.02189781e-01]\n",
      " [0.00000000e+00 3.20641283e-02 1.00200401e-02 3.20641283e-02\n",
      "  3.60721443e-02 7.33466934e-01 4.80961924e-02 2.40480962e-02\n",
      "  2.60521042e-02 6.01202405e-03 2.00400802e-03 5.01002004e-02]\n",
      " [0.00000000e+00 2.23880597e-02 4.97512438e-03 9.95024876e-03\n",
      "  1.49253731e-02 3.48258706e-02 8.73134328e-01 7.46268657e-03\n",
      "  0.00000000e+00 9.95024876e-03 1.49253731e-02 7.46268657e-03]\n",
      " [0.00000000e+00 7.34463277e-02 1.69491525e-02 1.88323917e-02\n",
      "  1.07344633e-01 1.88323917e-02 0.00000000e+00 6.94915254e-01\n",
      "  1.12994350e-02 0.00000000e+00 0.00000000e+00 5.83804143e-02]\n",
      " [0.00000000e+00 6.17283951e-02 1.23456790e-02 2.22222222e-01\n",
      "  7.40740741e-02 4.19753086e-02 0.00000000e+00 2.22222222e-02\n",
      "  4.93827160e-01 0.00000000e+00 0.00000000e+00 7.16049383e-02]\n",
      " [0.00000000e+00 0.00000000e+00 1.79211470e-03 0.00000000e+00\n",
      "  0.00000000e+00 0.00000000e+00 7.34767025e-02 0.00000000e+00\n",
      "  0.00000000e+00 9.24731183e-01 0.00000000e+00 0.00000000e+00]\n",
      " [1.63398693e-03 1.63398693e-03 0.00000000e+00 0.00000000e+00\n",
      "  4.90196078e-03 1.63398693e-03 3.26797386e-03 0.00000000e+00\n",
      "  1.63398693e-03 0.00000000e+00 9.85294118e-01 0.00000000e+00]\n",
      " [0.00000000e+00 5.18962076e-02 5.98802395e-03 1.59680639e-02\n",
      "  1.15768463e-01 2.19560878e-02 0.00000000e+00 3.59281437e-02\n",
      "  1.99600798e-02 0.00000000e+00 1.19760479e-02 7.20558882e-01]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1440 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "cm_xgb_bal_multi = confusion_matrix(bal_multi_y_test, bal_multi_xgb_preds)\n",
    "classes = balanced_df['Label'].unique()\n",
    "plot_confusion_matrix(cm_xgb_bal_multi, classes, normalize=True,\n",
    "                      title='XGBoost Confusion Matrix', cmap=plt.cm.Blues)\n",
    "\n",
    "plt.savefig('Images/bal_multi_xgb_conf_matrix.png', dpi = 100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Anomaly Detection Dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Binary Classification (Anomaly Detection)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train Test Split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [],
   "source": [
    "anom_binary_X = anomaly_df.drop(['Flow_ID', 'Label', 'HOPOPT', 'Malicious'], axis=1)\n",
    "anom_binary_y = anomaly_df['Malicious']\n",
    "\n",
    "anom_binary_X_train, anom_binary_X_test, anom_binary_y_train, anom_binary_y_test = train_test_split(\n",
    "    anom_binary_X, anom_binary_y, stratify=anom_binary_y, test_size=0.25, random_state=0)\n",
    "\n",
    "# Convert to list to avoid type error\n",
    "anom_binary_y_train = list(anom_binary_y_train)\n",
    "anom_binary_y_test = list(anom_binary_y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Standard Scalar"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [],
   "source": [
    "std_scaler = StandardScaler()\n",
    "\n",
    "anom_binary_X_train_scl = std_scaler.fit_transform(anom_binary_X_train)\n",
    "anom_binary_X_test_scl = std_scaler.transform(anom_binary_X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Dummy Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/eric/opt/anaconda3/lib/python3.7/site-packages/sklearn/dummy.py:132: FutureWarning: The default value of strategy will change from stratified to prior in 0.24.\n",
      "  \"stratified to prior in 0.24.\", FutureWarning)\n"
     ]
    }
   ],
   "source": [
    "anom_binary_dummy = DummyClassifier(random_state=0)\n",
    "anom_binary_dummy.fit(anom_binary_X_train_scl, anom_binary_y_train)\n",
    "anom_binary_dummy_preds = list(anom_binary_dummy.predict(anom_binary_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.9777091491113331\n",
      "      F1: 0.009102730819245775\n"
     ]
    }
   ],
   "source": [
    "anom_binary_dummy_acc = accuracy_score(anom_binary_y_test, anom_binary_dummy_preds)\n",
    "anom_binary_dummy_f1 = f1_score(anom_binary_y_test, anom_binary_dummy_preds)\n",
    "\n",
    "print('Accuracy:', anom_binary_dummy_acc)\n",
    "print('      F1:', anom_binary_dummy_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Gaussian Naive Bayes "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [],
   "source": [
    "anom_binary_gnb = GaussianNB()\n",
    "anom_binary_gnb.fit(anom_binary_X_train_scl, anom_binary_y_train)\n",
    "anom_binary_gnb_preds = list(anom_binary_gnb.predict(anom_binary_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.6333334726337388\n",
      "      F1: 0.055197811901018676\n"
     ]
    }
   ],
   "source": [
    "anom_binary_gnb_acc = accuracy_score(anom_binary_y_test, anom_binary_gnb_preds)\n",
    "anom_binary_gnb_f1 = f1_score(anom_binary_y_test, anom_binary_gnb_preds)\n",
    "\n",
    "print('Accuracy:', anom_binary_gnb_acc)\n",
    "print('      F1:', anom_binary_gnb_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Decision Tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [],
   "source": [
    "anom_binary_dt = DecisionTreeClassifier(random_state=0)\n",
    "anom_binary_dt.fit(anom_binary_X_train_scl, anom_binary_y_train)\n",
    "anom_binary_dt_preds = list(anom_binary_dt.predict(anom_binary_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.9998579135863864\n",
      "      F1: 0.9936802973977695\n"
     ]
    }
   ],
   "source": [
    "anom_binary_dt_acc = accuracy_score(anom_binary_y_test, anom_binary_dt_preds)\n",
    "anom_binary_dt_f1 = f1_score(anom_binary_y_test, anom_binary_dt_preds)\n",
    "\n",
    "print('Accuracy:', anom_binary_dt_acc)\n",
    "print('      F1:', anom_binary_dt_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### K Nearest Neighbors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "anom_binary_knn = KNeighborsClassifier()\n",
    "anom_binary_knn.fit(anom_binary_X_train_scl, anom_binary_y_train)\n",
    "anom_binary_knn_preds = list(anom_binary_knn.predict(anom_binary_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "anom_binary_knn_acc = accuracy_score(anom_binary_y_test, anom_binary_knn_preds)\n",
    "anom_binary_knn_f1 = f1_score(anom_binary_y_test, anom_binary_knn_preds)\n",
    "\n",
    "print('Accuracy:', anom_binary_knn_acc)\n",
    "print('      F1:', anom_binary_knn_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "anom_binary_rf = RandomForestClassifier(random_state=0, n_jobs=-1)\n",
    "anom_binary_rf.fit(anom_binary_X_train_scl, anom_binary_y_train)\n",
    "anom_binary_rf_preds = list(anom_binary_rf.predict(anom_binary_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.9988847583643122\n",
      "      F1: 0.9988954344624447\n"
     ]
    }
   ],
   "source": [
    "anom_binary_rf_acc = accuracy_score(anom_binary_y_test, anom_binary_rf_preds)\n",
    "anom_binary_rf_f1 = f1_score(anom_binary_y_test, anom_binary_rf_preds)\n",
    "\n",
    "print('Accuracy:', anom_binary_rf_acc)\n",
    "print('      F1:', anom_binary_rf_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### XGBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "anom_binary_xgb = XGBClassifier(random_state=0, n_jobs=-1)\n",
    "anom_binary_xgb.fit(anom_binary_X_train_scl, anom_binary_y_train)\n",
    "anom_binary_xgb_preds = list(anom_binary_xgb.predict(anom_binary_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.9994423791821562\n",
      "      F1: 0.9994477172312224\n"
     ]
    }
   ],
   "source": [
    "anom_binary_xgb_acc = accuracy_score(anom_binary_y_test, anom_binary_xgb_preds)\n",
    "anom_binary_xgb_f1 = f1_score(anom_binary_y_test, anom_binary_xgb_preds)\n",
    "\n",
    "print('Accuracy:', anom_binary_xgb_acc)\n",
    "print('      F1:', anom_binary_xgb_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature Importances"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Decision Tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('Bwd_Pkt_Length_Max', 0.4321718724387804),\n",
       " ('Fwd_IAT_Min', 0.09638106984906894),\n",
       " ('ACK_Flag_Count', 0.09033950805271436),\n",
       " ('Bwd_Pkts_Sec', 0.04781227881326841),\n",
       " ('Pkt_Size_Mean', 0.047542461074881436),\n",
       " ('Fwd_Pkt_Length_Std', 0.047491182791384416),\n",
       " ('Fwd_Init_Win_Bytes', 0.041885485142335525),\n",
       " ('Bwd_Header_Length', 0.03417186684692964),\n",
       " ('Bwd_Init_Win_Bytes', 0.029978998366472578),\n",
       " ('URG_Flag_Count', 0.028887571059214186)]"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anom_binary_dt_features = dict(zip(anom_binary_X.columns, anom_binary_dt.feature_importances_))\n",
    "anom_binary_dt_features = sorted(anom_binary_dt_features.items(), key=lambda x: x[1], reverse=True)\n",
    "anom_binary_dt_features[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('Bwd_Segment_Size_Mean', 0.06479943731326662),\n",
       " ('Bwd_Pkt_Length_Max', 0.05768429926659383),\n",
       " ('Fwd_Init_Win_Bytes', 0.051959549429500464),\n",
       " ('Bwd_Pkt_Length_Mean', 0.051723724983398243),\n",
       " ('Pkt_Length_Var', 0.05011734358632702),\n",
       " ('Pkt_Length_Min', 0.03474782546490374),\n",
       " ('Bwd_Subflow_Bytes', 0.03389031994987526),\n",
       " ('Fwd_Pkt_Length_Min', 0.031065625410276246),\n",
       " ('Bwd_Total_Bytes', 0.02808381988258205),\n",
       " ('Pkt_Length_Std', 0.026423405011124897)]"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anom_binary_rf_features = dict(zip(anom_binary_X.columns, anom_binary_rf.feature_importances_))\n",
    "anom_binary_rf_features = sorted(anom_binary_rf_features.items(), key=lambda x: x[1], reverse=True)\n",
    "anom_binary_rf_features[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "XGBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('Bwd_Pkt_Length_Max', 0.7066348),\n",
       " ('Bwd_Pkt_Length_Mean', 0.046133853),\n",
       " ('Fwd_Pkt_Length_Std', 0.02572131),\n",
       " ('ACK_Flag_Count', 0.023738889),\n",
       " ('Bwd_Pkts_Sec', 0.018046636),\n",
       " ('Bwd_Header_Length', 0.01647789),\n",
       " ('PSH_Flag_Count', 0.015690256),\n",
       " ('Fwd_IAT_Min', 0.015220775),\n",
       " ('Fwd_Total_Pkts', 0.015060951),\n",
       " ('Pkt_Size_Mean', 0.011617822)]"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anom_binary_xgb_features = dict(zip(anom_binary_X.columns, anom_binary_xgb.feature_importances_))\n",
    "anom_binary_xgb_features = sorted(anom_binary_xgb_features.items(), key=lambda x: x[1], reverse=True)\n",
    "anom_binary_xgb_features[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Multiclass Classification (Anomaly Detection)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train Test Split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "anom_multi_X = anomaly_df.drop(['Flow_ID', 'Label', 'HOPOPT', 'Malicious'], axis=1)\n",
    "anom_multi_y = anomaly_df['Label']\n",
    "\n",
    "anom_multi_X_train, anom_multi_X_test, anom_multi_y_train, anom_multi_y_test = train_test_split(\n",
    "    anom_multi_X, anom_multi_y, stratify=anom_multi_y, test_size=0.25, random_state=0)\n",
    "\n",
    "# Convert to list to avoid type error\n",
    "anom_multi_y_train = list(anom_multi_y_train)\n",
    "anom_multi_y_test = list(anom_multi_y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Standard Scalar"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "std_scaler = StandardScaler()\n",
    "\n",
    "anom_multi_X_train_scl = std_scaler.fit_transform(anom_multi_X_train)\n",
    "anom_multi_X_test_scl = std_scaler.transform(anom_multi_X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Dummy Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/eric/opt/anaconda3/lib/python3.7/site-packages/sklearn/dummy.py:132: FutureWarning: The default value of strategy will change from stratified to prior in 0.24.\n",
      "  \"stratified to prior in 0.24.\", FutureWarning)\n"
     ]
    }
   ],
   "source": [
    "anom_multi_dummy = DummyClassifier(random_state=0)\n",
    "anom_multi_dummy.fit(anom_multi_X_train_scl, anom_multi_y_train)\n",
    "anom_multi_dummy_preds = list(anom_multi_dummy.predict(anom_multi_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.2776951672862453\n",
      "      F1: 0.27377289188976395\n"
     ]
    }
   ],
   "source": [
    "anom_multi_dummy_acc = accuracy_score(anom_multi_y_test, anom_multi_dummy_preds)\n",
    "anom_multi_dummy_f1 = f1_score(anom_multi_y_test, anom_multi_dummy_preds, average='weighted')\n",
    "\n",
    "print('Accuracy:', anom_multi_dummy_acc)\n",
    "print('      F1:', anom_multi_dummy_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Gaussian Naive Bayes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "anom_multi_gnb = GaussianNB()\n",
    "anom_multi_gnb.fit(anom_multi_X_train_scl, anom_multi_y_train)\n",
    "anom_multi_gnb_preds = list(anom_multi_gnb.predict(anom_multi_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.3140334572490706\n",
      "      F1: 0.3481777387956166\n"
     ]
    }
   ],
   "source": [
    "anom_multi_gnb_acc = accuracy_score(anom_multi_y_test, anom_multi_gnb_preds)\n",
    "anom_multi_gnb_f1 = f1_score(anom_multi_y_test, anom_multi_gnb_preds, average='weighted')\n",
    "\n",
    "print('Accuracy:', anom_multi_gnb_acc)\n",
    "print('      F1:', anom_multi_gnb_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Decision Tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "anom_multi_dt = DecisionTreeClassifier(random_state=0)\n",
    "anom_multi_dt.fit(anom_multi_X_train_scl, anom_multi_y_train)\n",
    "anom_multi_dt_preds = list(anom_multi_dt.predict(anom_multi_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.8467472118959107\n",
      "      F1: 0.8472017627615264\n"
     ]
    }
   ],
   "source": [
    "anom_multi_dt_acc = accuracy_score(anom_multi_y_test, anom_multi_dt_preds)\n",
    "anom_multi_dt_f1 = f1_score(anom_multi_y_test, anom_multi_dt_preds, average='weighted')\n",
    "\n",
    "print('Accuracy:', anom_multi_dt_acc)\n",
    "print('      F1:', anom_multi_dt_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### K Nearest Neighbors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "anom_multi_knn = KNeighborsClassifier()\n",
    "anom_multi_knn.fit(anom_multi_X_train_scl, anom_multi_y_train)\n",
    "anom_multi_knn_preds = list(anom_multi_knn.predict(anom_multi_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.8433085501858736\n",
      "      F1: 0.8396681128702419\n"
     ]
    }
   ],
   "source": [
    "anom_multi_knn_acc = accuracy_score(anom_multi_y_test, anom_multi_knn_preds)\n",
    "anom_multi_knn_f1 = f1_score(anom_multi_y_test, anom_multi_knn_preds, average='weighted')\n",
    "\n",
    "print('Accuracy:', anom_multi_knn_acc)\n",
    "print('      F1:', anom_multi_knn_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "anom_multi_rf = RandomForestClassifier(random_state=0, n_jobs=-1)\n",
    "anom_multi_rf.fit(anom_multi_X_train_scl, anom_multi_y_train)\n",
    "anom_multi_rf_preds = list(anom_multi_rf.predict(anom_multi_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.8523234200743495\n",
      "      F1: 0.8519008118496785\n"
     ]
    }
   ],
   "source": [
    "anom_multi_rf_acc = accuracy_score(anom_multi_y_test, anom_multi_rf_preds)\n",
    "anom_multi_rf_f1 = f1_score(anom_multi_y_test, anom_multi_rf_preds, average='weighted')\n",
    "\n",
    "print('Accuracy:', anom_multi_rf_acc)\n",
    "print('      F1:', anom_multi_rf_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### XGBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "anom_multi_xgb = XGBClassifier(random_state=0, n_jobs=-1)\n",
    "anom_multi_xgb.fit(anom_multi_X_train_scl, anom_multi_y_train)\n",
    "anom_multi_xgb_preds = list(anom_multi_xgb.predict(anom_multi_X_test_scl))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.8650557620817844\n",
      "      F1: 0.8647730500675536\n"
     ]
    }
   ],
   "source": [
    "anom_multi_xgb_acc = accuracy_score(anom_multi_y_test, anom_multi_xgb_preds)\n",
    "anom_multi_xgb_f1 = f1_score(anom_multi_y_test, anom_multi_xgb_preds, average='weighted')\n",
    "\n",
    "print('Accuracy:', anom_multi_xgb_acc)\n",
    "print('      F1:', anom_multi_xgb_f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature Importances"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Decision Tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('Bwd_Pkt_Length_Max', 0.17813608910486856),\n",
       " ('Pkt_Length_Min', 0.13232021964005997),\n",
       " ('Fwd_Init_Win_Bytes', 0.07129227402814149),\n",
       " ('Flow_Bytes_Sec', 0.05254508982369269),\n",
       " ('ACK_Flag_Count', 0.04822122872673132),\n",
       " ('Fwd_IAT_Min', 0.048125155685704676),\n",
       " ('Flow_IAT_Max', 0.04057134077946873),\n",
       " ('Bwd_Pkts_Sec', 0.02851069734541092),\n",
       " ('Fwd_Subflow_Bytes', 0.027439768028378247),\n",
       " ('Bwd_Header_Length', 0.027045377146699245)]"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anom_multi_dt_features = dict(zip(anom_multi_X.columns, anom_multi_dt.feature_importances_))\n",
    "anom_multi_dt_features = sorted(anom_multi_dt_features.items(), key=lambda x: x[1], reverse=True)\n",
    "anom_multi_dt_features[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('Fwd_Init_Win_Bytes', 0.052314072135577656),\n",
       " ('Pkt_Size_Mean', 0.033927197406239375),\n",
       " ('Pkt_Length_Min', 0.03362226368219828),\n",
       " ('Flow_IAT_Mean', 0.03229548261837191),\n",
       " ('Fwd_Pkt_Length_Min', 0.02929883891176453),\n",
       " ('Fwd_Segment_Size_Mean', 0.029189954369181214),\n",
       " ('Pkt_Length_Mean', 0.02841634731314821),\n",
       " ('Bwd_Segment_Size_Mean', 0.027442983138788266),\n",
       " ('Fwd_Pkts_Sec', 0.02733519211706754),\n",
       " ('Flow_IAT_Max', 0.02732665349636803)]"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anom_multi_rf_features = dict(zip(anom_multi_X.columns, anom_multi_rf.feature_importances_))\n",
    "anom_multi_rf_features = sorted(anom_multi_rf_features.items(), key=lambda x: x[1], reverse=True)\n",
    "anom_multi_rf_features[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "XGBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('Bwd_Pkt_Length_Max', 0.8722749),\n",
       " ('Pkt_Length_Min', 0.014750869),\n",
       " ('ACK_Flag_Count', 0.01015347),\n",
       " ('CWE_Flag_Count', 0.008745772),\n",
       " ('Time_Idle_Std', 0.0071296976),\n",
       " ('PSH_Flag_Count', 0.005871469),\n",
       " ('Bwd_Header_Length', 0.0050045243),\n",
       " ('Fwd_Total_Bytes', 0.0049487054),\n",
       " ('UDP', 0.004383982),\n",
       " ('URG_Flag_Count', 0.0042051356)]"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anom_multi_xgb_features = dict(zip(anom_multi_X.columns, anom_multi_xgb.feature_importances_))\n",
    "anom_multi_xgb_features = sorted(anom_multi_xgb_features.items(), key=lambda x: x[1], reverse=True)\n",
    "anom_multi_xgb_features[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Confusion Matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Normalized confusion matrix\n",
      "[[9.99624765e-01 0.00000000e+00 1.87617261e-04 0.00000000e+00\n",
      "  0.00000000e+00 0.00000000e+00 1.87617261e-04 0.00000000e+00\n",
      "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
      " [1.70648464e-03 7.06484642e-01 2.21843003e-02 3.92491468e-02\n",
      "  6.14334471e-02 2.73037543e-02 8.53242321e-03 4.60750853e-02\n",
      "  4.09556314e-02 0.00000000e+00 1.70648464e-03 4.43686007e-02]\n",
      " [0.00000000e+00 6.50000000e-02 5.77500000e-01 2.50000000e-01\n",
      "  1.25000000e-02 2.00000000e-02 1.50000000e-02 3.25000000e-02\n",
      "  1.50000000e-02 5.00000000e-03 0.00000000e+00 7.50000000e-03]\n",
      " [0.00000000e+00 7.21649485e-02 3.32474227e-01 3.73711340e-01\n",
      "  6.44329897e-02 3.35051546e-02 2.83505155e-02 2.31958763e-02\n",
      "  3.35051546e-02 5.15463918e-03 0.00000000e+00 3.35051546e-02]\n",
      " [1.82481752e-03 4.19708029e-02 3.64963504e-03 7.29927007e-03\n",
      "  7.73722628e-01 5.47445255e-03 5.47445255e-03 4.74452555e-02\n",
      "  7.29927007e-03 0.00000000e+00 3.64963504e-03 1.02189781e-01]\n",
      " [0.00000000e+00 3.20641283e-02 1.00200401e-02 3.20641283e-02\n",
      "  3.60721443e-02 7.33466934e-01 4.80961924e-02 2.40480962e-02\n",
      "  2.60521042e-02 6.01202405e-03 2.00400802e-03 5.01002004e-02]\n",
      " [0.00000000e+00 2.23880597e-02 4.97512438e-03 9.95024876e-03\n",
      "  1.49253731e-02 3.48258706e-02 8.73134328e-01 7.46268657e-03\n",
      "  0.00000000e+00 9.95024876e-03 1.49253731e-02 7.46268657e-03]\n",
      " [0.00000000e+00 7.34463277e-02 1.69491525e-02 1.88323917e-02\n",
      "  1.07344633e-01 1.88323917e-02 0.00000000e+00 6.94915254e-01\n",
      "  1.12994350e-02 0.00000000e+00 0.00000000e+00 5.83804143e-02]\n",
      " [0.00000000e+00 6.17283951e-02 1.23456790e-02 2.22222222e-01\n",
      "  7.40740741e-02 4.19753086e-02 0.00000000e+00 2.22222222e-02\n",
      "  4.93827160e-01 0.00000000e+00 0.00000000e+00 7.16049383e-02]\n",
      " [0.00000000e+00 0.00000000e+00 1.79211470e-03 0.00000000e+00\n",
      "  0.00000000e+00 0.00000000e+00 7.34767025e-02 0.00000000e+00\n",
      "  0.00000000e+00 9.24731183e-01 0.00000000e+00 0.00000000e+00]\n",
      " [1.63398693e-03 1.63398693e-03 0.00000000e+00 0.00000000e+00\n",
      "  4.90196078e-03 1.63398693e-03 3.26797386e-03 0.00000000e+00\n",
      "  1.63398693e-03 0.00000000e+00 9.85294118e-01 0.00000000e+00]\n",
      " [0.00000000e+00 5.18962076e-02 5.98802395e-03 1.59680639e-02\n",
      "  1.15768463e-01 2.19560878e-02 0.00000000e+00 3.59281437e-02\n",
      "  1.99600798e-02 0.00000000e+00 1.19760479e-02 7.20558882e-01]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1440 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "cm_xgb_anom_multi = confusion_matrix(anom_multi_y_test, anom_multi_xgb_preds)\n",
    "classes = anomaly_df['Label'].unique()\n",
    "plot_confusion_matrix(cm_xgb_anom_multi, classes, normalize=True,\n",
    "                      title='XGBoost Confusion Matrix', cmap=plt.cm.Blues)\n",
    "\n",
    "# plt.savefig('Images/bal_multi_xgb_conf_matrix.png', dpi = 100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Export Model Results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "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>Score</th>\n",
       "      <th>Bal_Binary_Dummy</th>\n",
       "      <th>Bal_Binary_GNB</th>\n",
       "      <th>Bal_Binary_DT</th>\n",
       "      <th>Bal_Binary_KNN</th>\n",
       "      <th>Bal_Binary_RF</th>\n",
       "      <th>Bal_Binary_XGB</th>\n",
       "      <th>Bal_Multicalss_Dummy</th>\n",
       "      <th>Bal_Multiclass_GNB</th>\n",
       "      <th>Bal_Multiclass_DT</th>\n",
       "      <th>Bal_Multiclass_KNN</th>\n",
       "      <th>Bal_Multiclass_RF</th>\n",
       "      <th>Bal_Multiclass_XGB</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Accuracy</td>\n",
       "      <td>0.503996</td>\n",
       "      <td>0.694331</td>\n",
       "      <td>0.997955</td>\n",
       "      <td>0.996004</td>\n",
       "      <td>0.998885</td>\n",
       "      <td>0.999442</td>\n",
       "      <td>0.277695</td>\n",
       "      <td>0.314033</td>\n",
       "      <td>0.846747</td>\n",
       "      <td>0.843309</td>\n",
       "      <td>0.852323</td>\n",
       "      <td>0.865056</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>F1_Score</td>\n",
       "      <td>0.501075</td>\n",
       "      <td>0.749600</td>\n",
       "      <td>0.997975</td>\n",
       "      <td>0.996042</td>\n",
       "      <td>0.998895</td>\n",
       "      <td>0.999448</td>\n",
       "      <td>0.273773</td>\n",
       "      <td>0.749600</td>\n",
       "      <td>0.847202</td>\n",
       "      <td>0.839668</td>\n",
       "      <td>0.851901</td>\n",
       "      <td>0.864773</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Score  Bal_Binary_Dummy  Bal_Binary_GNB  Bal_Binary_DT  Bal_Binary_KNN  \\\n",
       "0  Accuracy          0.503996        0.694331       0.997955        0.996004   \n",
       "1  F1_Score          0.501075        0.749600       0.997975        0.996042   \n",
       "\n",
       "   Bal_Binary_RF  Bal_Binary_XGB  Bal_Multicalss_Dummy  Bal_Multiclass_GNB  \\\n",
       "0       0.998885        0.999442              0.277695            0.314033   \n",
       "1       0.998895        0.999448              0.273773            0.749600   \n",
       "\n",
       "   Bal_Multiclass_DT  Bal_Multiclass_KNN  Bal_Multiclass_RF  \\\n",
       "0           0.846747            0.843309           0.852323   \n",
       "1           0.847202            0.839668           0.851901   \n",
       "\n",
       "   Bal_Multiclass_XGB  \n",
       "0            0.865056  \n",
       "1            0.864773  "
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_scores = pd.DataFrame({\n",
    "    'Score':['Accuracy', 'F1_Score'],\n",
    "    'Bal_Binary_Dummy':[bal_binary_dummy_acc, bal_binary_dummy_f1],\n",
    "    'Bal_Binary_GNB':[bal_binary_gnb_acc, bal_binary_gnb_f1],\n",
    "    'Bal_Binary_DT':[bal_binary_dt_acc, bal_binary_dt_f1],\n",
    "    'Bal_Binary_KNN':[bal_binary_knn_acc, bal_binary_knn_f1],\n",
    "    'Bal_Binary_RF':[bal_binary_rf_acc, bal_binary_rf_f1],\n",
    "    'Bal_Binary_XGB':[bal_binary_xgb_acc, bal_binary_xgb_f1],\n",
    "    'Bal_Multicalss_Dummy':[bal_multi_dummy_acc, bal_multi_dummy_f1],\n",
    "    'Bal_Multiclass_GNB':[bal_multi_gnb_acc, bal_binary_gnb_f1],\n",
    "    'Bal_Multiclass_DT':[bal_multi_dt_acc, bal_multi_dt_f1],\n",
    "    'Bal_Multiclass_KNN':[bal_multi_knn_acc, bal_multi_knn_f1],\n",
    "    'Bal_Multiclass_RF':[bal_multi_rf_acc, bal_multi_rf_f1],\n",
    "    'Bal_Multiclass_XGB':[bal_multi_xgb_acc, bal_multi_xgb_f1]\n",
    "})\n",
    "\n",
    "model_scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [],
   "source": [
    "# model_scores.to_csv('Model_Results/Model_Scores.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Feature importances"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [],
   "source": [
    "features = []\n",
    "for i in range(1, len(bal_binary_dt_features)+1):\n",
    "    features.append(f'Feature_{i}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "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>Feature</th>\n",
       "      <th>Bal_Binary_DT</th>\n",
       "      <th>Bal_Binary_RF</th>\n",
       "      <th>Bal_Binary_XGB</th>\n",
       "      <th>Bal_Multiclass_DT</th>\n",
       "      <th>Bal_Multiclass_RF</th>\n",
       "      <th>Bal_Multiclass_XGB</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Feature_1</td>\n",
       "      <td>(Bwd_Pkt_Length_Max, 0.4321718724387804)</td>\n",
       "      <td>(Bwd_Segment_Size_Mean, 0.06479943731326662)</td>\n",
       "      <td>(Bwd_Pkt_Length_Max, 0.7066348)</td>\n",
       "      <td>(Bwd_Pkt_Length_Max, 0.17813608910486856)</td>\n",
       "      <td>(Fwd_Init_Win_Bytes, 0.052314072135577656)</td>\n",
       "      <td>(Bwd_Pkt_Length_Max, 0.8722749)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Feature_2</td>\n",
       "      <td>(Fwd_IAT_Min, 0.09638106984906894)</td>\n",
       "      <td>(Bwd_Pkt_Length_Max, 0.05768429926659383)</td>\n",
       "      <td>(Bwd_Pkt_Length_Mean, 0.046133853)</td>\n",
       "      <td>(Pkt_Length_Min, 0.13232021964005997)</td>\n",
       "      <td>(Pkt_Size_Mean, 0.033927197406239375)</td>\n",
       "      <td>(Pkt_Length_Min, 0.014750869)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Feature_3</td>\n",
       "      <td>(ACK_Flag_Count, 0.09033950805271436)</td>\n",
       "      <td>(Fwd_Init_Win_Bytes, 0.051959549429500464)</td>\n",
       "      <td>(Fwd_Pkt_Length_Std, 0.02572131)</td>\n",
       "      <td>(Fwd_Init_Win_Bytes, 0.07129227402814149)</td>\n",
       "      <td>(Pkt_Length_Min, 0.03362226368219828)</td>\n",
       "      <td>(ACK_Flag_Count, 0.01015347)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Feature_4</td>\n",
       "      <td>(Bwd_Pkts_Sec, 0.04781227881326841)</td>\n",
       "      <td>(Bwd_Pkt_Length_Mean, 0.051723724983398243)</td>\n",
       "      <td>(ACK_Flag_Count, 0.023738889)</td>\n",
       "      <td>(Flow_Bytes_Sec, 0.05254508982369269)</td>\n",
       "      <td>(Flow_IAT_Mean, 0.03229548261837191)</td>\n",
       "      <td>(CWE_Flag_Count, 0.008745772)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Feature_5</td>\n",
       "      <td>(Pkt_Size_Mean, 0.047542461074881436)</td>\n",
       "      <td>(Pkt_Length_Var, 0.05011734358632702)</td>\n",
       "      <td>(Bwd_Pkts_Sec, 0.018046636)</td>\n",
       "      <td>(ACK_Flag_Count, 0.04822122872673132)</td>\n",
       "      <td>(Fwd_Pkt_Length_Min, 0.02929883891176453)</td>\n",
       "      <td>(Time_Idle_Std, 0.0071296976)</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Feature                             Bal_Binary_DT  \\\n",
       "0  Feature_1  (Bwd_Pkt_Length_Max, 0.4321718724387804)   \n",
       "1  Feature_2        (Fwd_IAT_Min, 0.09638106984906894)   \n",
       "2  Feature_3     (ACK_Flag_Count, 0.09033950805271436)   \n",
       "3  Feature_4       (Bwd_Pkts_Sec, 0.04781227881326841)   \n",
       "4  Feature_5     (Pkt_Size_Mean, 0.047542461074881436)   \n",
       "\n",
       "                                  Bal_Binary_RF  \\\n",
       "0  (Bwd_Segment_Size_Mean, 0.06479943731326662)   \n",
       "1     (Bwd_Pkt_Length_Max, 0.05768429926659383)   \n",
       "2    (Fwd_Init_Win_Bytes, 0.051959549429500464)   \n",
       "3   (Bwd_Pkt_Length_Mean, 0.051723724983398243)   \n",
       "4         (Pkt_Length_Var, 0.05011734358632702)   \n",
       "\n",
       "                       Bal_Binary_XGB  \\\n",
       "0     (Bwd_Pkt_Length_Max, 0.7066348)   \n",
       "1  (Bwd_Pkt_Length_Mean, 0.046133853)   \n",
       "2    (Fwd_Pkt_Length_Std, 0.02572131)   \n",
       "3       (ACK_Flag_Count, 0.023738889)   \n",
       "4         (Bwd_Pkts_Sec, 0.018046636)   \n",
       "\n",
       "                           Bal_Multiclass_DT  \\\n",
       "0  (Bwd_Pkt_Length_Max, 0.17813608910486856)   \n",
       "1      (Pkt_Length_Min, 0.13232021964005997)   \n",
       "2  (Fwd_Init_Win_Bytes, 0.07129227402814149)   \n",
       "3      (Flow_Bytes_Sec, 0.05254508982369269)   \n",
       "4      (ACK_Flag_Count, 0.04822122872673132)   \n",
       "\n",
       "                            Bal_Multiclass_RF               Bal_Multiclass_XGB  \n",
       "0  (Fwd_Init_Win_Bytes, 0.052314072135577656)  (Bwd_Pkt_Length_Max, 0.8722749)  \n",
       "1       (Pkt_Size_Mean, 0.033927197406239375)    (Pkt_Length_Min, 0.014750869)  \n",
       "2       (Pkt_Length_Min, 0.03362226368219828)     (ACK_Flag_Count, 0.01015347)  \n",
       "3        (Flow_IAT_Mean, 0.03229548261837191)    (CWE_Flag_Count, 0.008745772)  \n",
       "4   (Fwd_Pkt_Length_Min, 0.02929883891176453)    (Time_Idle_Std, 0.0071296976)  "
      ]
     },
     "execution_count": 102,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "feature_importances = pd.DataFrame({\n",
    "    'Feature':features,\n",
    "    'Bal_Binary_DT':bal_binary_dt_features,\n",
    "    'Bal_Binary_RF':bal_binary_rf_features,\n",
    "    'Bal_Binary_XGB':bal_binary_xgb_features,\n",
    "    'Bal_Multiclass_DT':bal_multi_dt_features,\n",
    "    'Bal_Multiclass_RF':bal_multi_rf_features,\n",
    "    'Bal_Multiclass_XGB':bal_multi_xgb_features\n",
    "})\n",
    "\n",
    "feature_importances.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [],
   "source": [
    "# feature_importances.to_csv('Model_Results/Feature_Importances.csv')"
   ]
  },
  {
   "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": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}