Performance Evaluation of Supervised Machine Learning Algorithms for Intrusion Detection.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Performance Evaluation of Supervised Machine Learning Algorithms for Intrusion Detection\n",
    "Data Mining and Predictive Analysis Lab Project\n",
    "Team Members:\n",
    "- Bhavya Sharma (170953130)\n",
    "- Raj Tulluri (170953084)\n",
    "- Ansh Raj (170953092)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Objective\n",
    "Intrusion detection system plays an important role in network security. Intrusion detection model is a predictive model used to predict the network data traffic as normal or intrusion. \n",
    "\n",
    "The objective of this project is to implement [this](https://www.researchgate.net/publication/306361859_Performance_Evaluation_of_Supervised_Machine_Learning_Algorithms_for_Intrusion_Detection) paper wherein classification and predictive models for intrusion detection are built by using machine learning classification algorithms namely Logistic Regression, Gaussian Naive Bayes, Support Vector Machine and Random Forest. We have added 2 more algorithms namely Gradient Boosting and Multilayer Perceptron. These algorithms are tested on the [NSL-KDD](https://www.kaggle.com/hassan06/nslkdd/kernels) data set.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## About the NSL-KDD data set\n",
    "Intrusion Detection Systems (IDS) are trained on internet traffic record data. The most common data set is the NSL-KDD, and is the benchmark for modern-day internet traffic.\n",
    "\n",
    "The NSL-KDD data set has the following advantages over the original KDD data set:\n",
    "- It does not include redundant records in the train set, so the classifiers will not be biased towards more frequent records.\n",
    "- There is no duplicate records in the proposed test sets; therefore, the performance of the learners are not biased by the methods which have better detection rates on the frequent records.\n",
    "- The number of selected records from each difficultylevel group is inversely proportional to the percentage of records in the original KDD data set. As a result, the classification rates of distinct machine learning methods vary in a wider range, which makes it more efficient to have an accurate evaluation of different learning techniques.\n",
    "- The number of records in the train and test sets are reasonable, which makes it affordable to run the experiments on the complete set without the need to randomly select a small portion. Consequently, evaluation results of different research works will be consistent and comparable.\n",
    "\n",
    "### Data files\n",
    "\n",
    "- KDDTrain+.ARFF: The full NSL-KDD train set with binary labels in ARFF format\n",
    "- KDDTrain+.TXT: The full NSL-KDD train set including attack-type labels and difficulty level in CSV format\n",
    "- KDDTrain+_20Percent.ARFF: A 20% subset of the KDDTrain+.arff file\n",
    "- KDDTrain+_20Percent.TXT: A 20% subset of the KDDTrain+.txt file\n",
    "- KDDTest+.ARFF: The full NSL-KDD test set with binary labels in ARFF format\n",
    "- KDDTest+.TXT: The full NSL-KDD test set including attack-type labels and difficulty level in CSV format\n",
    "- KDDTest-21.ARFF: A subset of the KDDTest+.arff file which does not include records with difficulty level of 21 out of 21\n",
    "- KDDTest-21.TXT: A subset of the KDDTest+.txt file which does not include records with difficulty level of 21 out of 21\n",
    "\n",
    "We will be using KDDTrain+ as train and KDDTest+ as test data sets."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Attacks\n",
    "The NSL-KDD data set contains a total of 39 attacks which can be classified into 4 categories: Denial of Service (DoS), Probe, User to Root(U2R), and Remote to Local (R2L). These attacks are identified based on the combination of the 41 features available. A brief description of each category can be seen below:\n",
    "\n",
    "1. DoS is an attack that tries to shut down traffic flow to and from the target system. The IDS is flooded with an abnormal amount of traffic, which the system can’t handle, and shuts down to protect itself. This prevents normal traffic from visiting a network. An example of this could be an online retailer getting flooded with online orders on a day with a big sale, and because the network can’t handle all the requests, it will shut down preventing paying customers to purchase anything. This is the most common attack in the data set. Examples: Neptune, Teardrop and Smurf.\n",
    "\n",
    "\n",
    "2. Probe or surveillance is an attack that tries to get information from a network. The goal here is to act like a thief and steal important information, whether it be personal information about clients or banking information. Examples: ipsweep, nmap attacks.\n",
    "\n",
    "\n",
    "3. U2R is an attack that starts off with a normal user account and tries to gain access to the system or network, as a super-user (root). The attacker attempts to exploit the vulnerabilities in a system to gain root privileges/access. Examples: Perl, Load Module and Eject attacks.\n",
    "\n",
    "\n",
    "4. R2L is an attack that tries to gain local access to a remote machine. An attacker does not have local access to the system/network, and tries to “hack” their way into the network. Examples: imap, guess password and ftp-write attacks.\n",
    "\n",
    "It is noticed from the descriptions above that DoS acts differently from the other three attacks, where DoS attempts to shut down a system to stop traffic flow altogether, whereas the other three attempts to quietly infiltrate the system undetected."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Reading the data sets and Initial Exploration\n",
    "Let's begin with importing the necessary modules."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.preprocessing import LabelEncoder, StandardScaler\n",
    "from sklearn.model_selection import train_test_split, cross_validate\n",
    "from sklearn.metrics import classification_report, confusion_matrix, precision_score, recall_score, f1_score\n",
    "from sklearn.metrics import roc_auc_score, roc_curve\n",
    "from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.neural_network import MLPClassifier\n",
    "from sklearn.calibration import calibration_curve\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.lines as mlines\n",
    "\n",
    "from warnings import filterwarnings\n",
    "filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note that the '.txt' files are stored in CSV format."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>33</th>\n",
       "      <th>34</th>\n",
       "      <th>35</th>\n",
       "      <th>36</th>\n",
       "      <th>37</th>\n",
       "      <th>38</th>\n",
       "      <th>39</th>\n",
       "      <th>40</th>\n",
       "      <th>41</th>\n",
       "      <th>42</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>tcp</td>\n",
       "      <td>ftp_data</td>\n",
       "      <td>SF</td>\n",
       "      <td>491</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.05</td>\n",
       "      <td>0.00</td>\n",
       "      <td>normal</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>udp</td>\n",
       "      <td>other</td>\n",
       "      <td>SF</td>\n",
       "      <td>146</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.60</td>\n",
       "      <td>0.88</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>normal</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>tcp</td>\n",
       "      <td>private</td>\n",
       "      <td>S0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.05</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>neptune</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>tcp</td>\n",
       "      <td>http</td>\n",
       "      <td>SF</td>\n",
       "      <td>232</td>\n",
       "      <td>8153</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>normal</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>tcp</td>\n",
       "      <td>http</td>\n",
       "      <td>SF</td>\n",
       "      <td>199</td>\n",
       "      <td>420</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>normal</td>\n",
       "      <td>21</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",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125968</th>\n",
       "      <td>0</td>\n",
       "      <td>tcp</td>\n",
       "      <td>private</td>\n",
       "      <td>S0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>neptune</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125969</th>\n",
       "      <td>8</td>\n",
       "      <td>udp</td>\n",
       "      <td>private</td>\n",
       "      <td>SF</td>\n",
       "      <td>105</td>\n",
       "      <td>145</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.96</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>normal</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125970</th>\n",
       "      <td>0</td>\n",
       "      <td>tcp</td>\n",
       "      <td>smtp</td>\n",
       "      <td>SF</td>\n",
       "      <td>2231</td>\n",
       "      <td>384</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.12</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.72</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>normal</td>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125971</th>\n",
       "      <td>0</td>\n",
       "      <td>tcp</td>\n",
       "      <td>klogin</td>\n",
       "      <td>S0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.05</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>neptune</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125972</th>\n",
       "      <td>0</td>\n",
       "      <td>tcp</td>\n",
       "      <td>ftp_data</td>\n",
       "      <td>SF</td>\n",
       "      <td>151</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.30</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.30</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>normal</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>125973 rows × 43 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        0    1         2   3     4     5   6   7   8   9   ...    33    34  \\\n",
       "0        0  tcp  ftp_data  SF   491     0   0   0   0   0  ...  0.17  0.03   \n",
       "1        0  udp     other  SF   146     0   0   0   0   0  ...  0.00  0.60   \n",
       "2        0  tcp   private  S0     0     0   0   0   0   0  ...  0.10  0.05   \n",
       "3        0  tcp      http  SF   232  8153   0   0   0   0  ...  1.00  0.00   \n",
       "4        0  tcp      http  SF   199   420   0   0   0   0  ...  1.00  0.00   \n",
       "...     ..  ...       ...  ..   ...   ...  ..  ..  ..  ..  ...   ...   ...   \n",
       "125968   0  tcp   private  S0     0     0   0   0   0   0  ...  0.10  0.06   \n",
       "125969   8  udp   private  SF   105   145   0   0   0   0  ...  0.96  0.01   \n",
       "125970   0  tcp      smtp  SF  2231   384   0   0   0   0  ...  0.12  0.06   \n",
       "125971   0  tcp    klogin  S0     0     0   0   0   0   0  ...  0.03  0.05   \n",
       "125972   0  tcp  ftp_data  SF   151     0   0   0   0   0  ...  0.30  0.03   \n",
       "\n",
       "          35    36    37    38    39    40       41  42  \n",
       "0       0.17  0.00  0.00  0.00  0.05  0.00   normal  20  \n",
       "1       0.88  0.00  0.00  0.00  0.00  0.00   normal  15  \n",
       "2       0.00  0.00  1.00  1.00  0.00  0.00  neptune  19  \n",
       "3       0.03  0.04  0.03  0.01  0.00  0.01   normal  21  \n",
       "4       0.00  0.00  0.00  0.00  0.00  0.00   normal  21  \n",
       "...      ...   ...   ...   ...   ...   ...      ...  ..  \n",
       "125968  0.00  0.00  1.00  1.00  0.00  0.00  neptune  20  \n",
       "125969  0.01  0.00  0.00  0.00  0.00  0.00   normal  21  \n",
       "125970  0.00  0.00  0.72  0.00  0.01  0.00   normal  18  \n",
       "125971  0.00  0.00  1.00  1.00  0.00  0.00  neptune  20  \n",
       "125972  0.30  0.00  0.00  0.00  0.00  0.00   normal  21  \n",
       "\n",
       "[125973 rows x 43 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train = pd.read_csv('KDDTrain+.txt', header=None)\n",
    "test = pd.read_csv('KDDTest+.txt', header=None)\n",
    "train"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "These data sets contain the records of the internet traffic seen by a simple intrusion detection network and are the ghosts of the traffic encountered by a real IDS and just the traces of its existence remains. The data set contains 43 features per record, with 41 of the features referring to the traffic input itself and the last two are labels (whether it is a normal or attack) and Score (the severity of the traffic input itself).\n",
    "\n",
    "A description of each feature and a breakdown of the data set can be seen in the google spreadsheet [here](https://docs.google.com/spreadsheets/d/1oAx320Vo9Z6HrBrL6BcfLH6sh2zIk9EKCv2OlaMGmwY/edit#gid=0)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's explore the label (column index 41)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "normal             67343\n",
       "neptune            41214\n",
       "satan               3633\n",
       "ipsweep             3599\n",
       "portsweep           2931\n",
       "smurf               2646\n",
       "nmap                1493\n",
       "back                 956\n",
       "teardrop             892\n",
       "warezclient          890\n",
       "pod                  201\n",
       "guess_passwd          53\n",
       "buffer_overflow       30\n",
       "warezmaster           20\n",
       "land                  18\n",
       "imap                  11\n",
       "rootkit               10\n",
       "loadmodule             9\n",
       "ftp_write              8\n",
       "multihop               7\n",
       "phf                    4\n",
       "perl                   3\n",
       "spy                    2\n",
       "Name: 41, dtype: int64"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.iloc[:,41].value_counts(dropna=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "normal             9711\n",
       "neptune            4657\n",
       "guess_passwd       1231\n",
       "mscan               996\n",
       "warezmaster         944\n",
       "apache2             737\n",
       "satan               735\n",
       "processtable        685\n",
       "smurf               665\n",
       "back                359\n",
       "snmpguess           331\n",
       "saint               319\n",
       "mailbomb            293\n",
       "snmpgetattack       178\n",
       "portsweep           157\n",
       "ipsweep             141\n",
       "httptunnel          133\n",
       "nmap                 73\n",
       "pod                  41\n",
       "buffer_overflow      20\n",
       "multihop             18\n",
       "named                17\n",
       "ps                   15\n",
       "sendmail             14\n",
       "rootkit              13\n",
       "xterm                13\n",
       "teardrop             12\n",
       "xlock                 9\n",
       "land                  7\n",
       "xsnoop                4\n",
       "ftp_write             3\n",
       "worm                  2\n",
       "udpstorm              2\n",
       "perl                  2\n",
       "sqlattack             2\n",
       "phf                   2\n",
       "loadmodule            2\n",
       "imap                  1\n",
       "Name: 41, dtype: int64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test.iloc[:,41].value_counts(dropna=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can observe that the test data set has more kinds of attacks than the train data set."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Cleaning and Visualization\n",
    "Our aim here is to check whether an intrusion is detected or not. We do not go deeper into the classification of the attacks. Thus, we add a 'target' column which maps label 'normal' to 0 and any intrusion as 1. We then plot a pie chart to visualize this distribution."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    67343\n",
       "1    58630\n",
       "Name: target, dtype: int64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train['target'] = train.iloc[:,41].apply(lambda x: 0 if x == 'normal' else 1)\n",
    "test['target'] = test.iloc[:,41].apply(lambda x: 0 if x == 'normal' else 1)\n",
    "data = train['target'].value_counts()\n",
    "lab = ['normal','attack']\n",
    "data "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "fig, ax = plt.subplots(figsize=(8,8), subplot_kw=dict(aspect=\"equal\"))\n",
    "def func(pct, allvals):\n",
    "    absolute = int(pct/100.*np.sum(allvals))\n",
    "    return \"{:.1f}%\\n({:d})\".format(pct, absolute)\n",
    "\n",
    "wedges, texts, autotexts = ax.pie(data, autopct=lambda pct: func(pct, data),\n",
    "                                  textprops=dict(color=\"w\"))\n",
    "ax.legend(wedges, lab,\n",
    "          title=\"Label\",\n",
    "          loc=\"center left\",\n",
    "          bbox_to_anchor=(1, 0, 0.5, 1))\n",
    "\n",
    "plt.setp(autotexts, size=10, weight=\"bold\")\n",
    "\n",
    "ax.set_title(\"Distribution of Labels\")\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we move on to the distribution of the 4 classes of attacks."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'DOS': 45927, 'Probe': 11656, 'U2R': 52, 'R2L': 85}"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "DOS = ['apache2','back','land','neptune','mailbomb','pod','processtable','smurf','teardrop','upstorm','worm']\n",
    "Probe = ['ipsweep','nmap','mscan','portsweep','saint','satan']\n",
    "U2R = ['buffer_overflow','loadmodule','perl','ps','rootkit','sqlattack','xterm']\n",
    "R2L = ['ftp_write','guess_passwd','httptunnel','imap','multihop','named','phf','sendmail','Snmpgetattack','spy','snmpguess','warzclient','warzmaster','xlock','xsnoop']\n",
    "count = {'DOS':0, 'Probe':0, 'U2R':0, 'R2L':0}\n",
    "for attack in train[41]:\n",
    "    if attack in DOS:\n",
    "        count['DOS'] += 1\n",
    "    elif attack in Probe:\n",
    "        count['Probe'] += 1\n",
    "    elif attack in U2R:\n",
    "        count['U2R'] += 1\n",
    "    elif attack in R2L:\n",
    "        count['R2L'] += 1\n",
    "count"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "[45927, 11656, 52, 85]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = list(count.values())\n",
    "lab = list(count.keys())\n",
    "fig, ax = plt.subplots(figsize=(8,8), subplot_kw=dict(aspect=\"equal\"))\n",
    "def func(pct, allvals):\n",
    "    absolute = int(pct/100.*np.sum(allvals))\n",
    "    return \"{:.1f}%\\n({:d})\".format(pct, absolute)\n",
    "\n",
    "wedges, texts, autotexts = ax.pie(data, autopct=lambda pct: func(pct, data),\n",
    "                                  textprops=dict(color=\"w\"))\n",
    "ax.legend(wedges, lab,\n",
    "          title=\"Label\",\n",
    "          loc=\"center left\",\n",
    "          bbox_to_anchor=(1, 0, 0.5, 1))\n",
    "\n",
    "plt.setp(autotexts, size=10, weight=\"bold\")\n",
    "\n",
    "ax.set_title(\"Distribution of Attacks\")\n",
    "\n",
    "plt.show()\n",
    "data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Column 42, specifies the severity of the attack. For the intrusion detection system, the severity of an attack is irrelevant as any attack can be sever. It does not help in classifying whether there was an attack or not, or which type of attack was observed. Thus the column 42 is removed from the data.<br>\n",
    "\n",
    "Initially, the model is supposed to detect whether there has been an attack or not (detection). Only then should it learn to classify the attacks. Thus the column 41 is removed for the initial model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": false
   },
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>32</th>\n",
       "      <th>33</th>\n",
       "      <th>34</th>\n",
       "      <th>35</th>\n",
       "      <th>36</th>\n",
       "      <th>37</th>\n",
       "      <th>38</th>\n",
       "      <th>39</th>\n",
       "      <th>40</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>tcp</td>\n",
       "      <td>ftp_data</td>\n",
       "      <td>SF</td>\n",
       "      <td>491</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>25</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.05</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>udp</td>\n",
       "      <td>other</td>\n",
       "      <td>SF</td>\n",
       "      <td>146</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.60</td>\n",
       "      <td>0.88</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>tcp</td>\n",
       "      <td>private</td>\n",
       "      <td>S0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>26</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.05</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>tcp</td>\n",
       "      <td>http</td>\n",
       "      <td>SF</td>\n",
       "      <td>232</td>\n",
       "      <td>8153</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>255</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>tcp</td>\n",
       "      <td>http</td>\n",
       "      <td>SF</td>\n",
       "      <td>199</td>\n",
       "      <td>420</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>255</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</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",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125968</th>\n",
       "      <td>0</td>\n",
       "      <td>tcp</td>\n",
       "      <td>private</td>\n",
       "      <td>S0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>25</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125969</th>\n",
       "      <td>8</td>\n",
       "      <td>udp</td>\n",
       "      <td>private</td>\n",
       "      <td>SF</td>\n",
       "      <td>105</td>\n",
       "      <td>145</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>244</td>\n",
       "      <td>0.96</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125970</th>\n",
       "      <td>0</td>\n",
       "      <td>tcp</td>\n",
       "      <td>smtp</td>\n",
       "      <td>SF</td>\n",
       "      <td>2231</td>\n",
       "      <td>384</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>30</td>\n",
       "      <td>0.12</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.72</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125971</th>\n",
       "      <td>0</td>\n",
       "      <td>tcp</td>\n",
       "      <td>klogin</td>\n",
       "      <td>S0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>8</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.05</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125972</th>\n",
       "      <td>0</td>\n",
       "      <td>tcp</td>\n",
       "      <td>ftp_data</td>\n",
       "      <td>SF</td>\n",
       "      <td>151</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>77</td>\n",
       "      <td>0.30</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.30</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>125973 rows × 42 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        0    1         2   3     4     5  6  7  8  9  ...   32    33    34  \\\n",
       "0       0  tcp  ftp_data  SF   491     0  0  0  0  0  ...   25  0.17  0.03   \n",
       "1       0  udp     other  SF   146     0  0  0  0  0  ...    1  0.00  0.60   \n",
       "2       0  tcp   private  S0     0     0  0  0  0  0  ...   26  0.10  0.05   \n",
       "3       0  tcp      http  SF   232  8153  0  0  0  0  ...  255  1.00  0.00   \n",
       "4       0  tcp      http  SF   199   420  0  0  0  0  ...  255  1.00  0.00   \n",
       "...    ..  ...       ...  ..   ...   ... .. .. .. ..  ...  ...   ...   ...   \n",
       "125968  0  tcp   private  S0     0     0  0  0  0  0  ...   25  0.10  0.06   \n",
       "125969  8  udp   private  SF   105   145  0  0  0  0  ...  244  0.96  0.01   \n",
       "125970  0  tcp      smtp  SF  2231   384  0  0  0  0  ...   30  0.12  0.06   \n",
       "125971  0  tcp    klogin  S0     0     0  0  0  0  0  ...    8  0.03  0.05   \n",
       "125972  0  tcp  ftp_data  SF   151     0  0  0  0  0  ...   77  0.30  0.03   \n",
       "\n",
       "          35    36    37    38    39    40  target  \n",
       "0       0.17  0.00  0.00  0.00  0.05  0.00       0  \n",
       "1       0.88  0.00  0.00  0.00  0.00  0.00       0  \n",
       "2       0.00  0.00  1.00  1.00  0.00  0.00       1  \n",
       "3       0.03  0.04  0.03  0.01  0.00  0.01       0  \n",
       "4       0.00  0.00  0.00  0.00  0.00  0.00       0  \n",
       "...      ...   ...   ...   ...   ...   ...     ...  \n",
       "125968  0.00  0.00  1.00  1.00  0.00  0.00       1  \n",
       "125969  0.01  0.00  0.00  0.00  0.00  0.00       0  \n",
       "125970  0.00  0.00  0.72  0.00  0.01  0.00       0  \n",
       "125971  0.00  0.00  1.00  1.00  0.00  0.00       1  \n",
       "125972  0.30  0.00  0.00  0.00  0.00  0.00       0  \n",
       "\n",
       "[125973 rows x 42 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.drop(columns=[41,42], inplace=True)\n",
    "test.drop(columns=[41,42], inplace=True)\n",
    "train"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "All the columns that are textual i.e. strings, are encoded to numerics for the understanding of the model. Since all the models applied cannot inherently handle textual data. A `LabelEncoder` is fit on the train data. The same encodings are used for the test data as well to maintain uniformity."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "categorical = [1,2,3]\n",
    "\n",
    "encoder = LabelEncoder()\n",
    "for col in categorical:\n",
    "    train.iloc[:,col] = pd.Series(encoder.fit_transform(train.iloc[:,col]))\n",
    "    test.iloc[:,col] = pd.Series(encoder.transform(test.iloc[:,col]))\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As shown previously, the test set contains types of attacks which are not in the train set. Every attack has a different kind of signature. The model can learn the signature of a type of an attack and can thus detect it in future. But for those attacks whose signatures are not known might not always be detected by the model. To avoid such an issue, the two sets are combined and then randomly split into train and test set again.<br>\n",
    "\n",
    "The train set will be used to train the model and test set for testing the model on unseen data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = pd.concat([train,test],axis=0)\n",
    "\n",
    "X = train.drop(columns='target',axis=1)\n",
    "y = train.target\n",
    "\n",
    "scaler = StandardScaler()\n",
    "scaled_X = pd.DataFrame(scaler.fit_transform(X))\n",
    "\n",
    "X_train,X_test,y_train,y_test = train_test_split(scaled_X,y,random_state=10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The following are the models which will be trained and compared for the intrusion detection problem statement.\n",
    "    \n",
    "  * Random Forest classifier\n",
    "  * Support Vector Machines\n",
    "  * Logistic Regression\n",
    "  * Gaussian Naive Bayes\n",
    "  * Gradient Boosting classifier\n",
    "  * Multi-Layer Perceptron classifier\n",
    "  \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Random Forest\n",
    "\n",
    "The Random Forest model, an ensemble technique, tackles the bias and overfitting a Decision Tree classifier is exposed to by fitting a forest (n_estimators i.e. large number) of Decision trees on the data, each fit on a random subset of data. It then aggregates the votes and outputs the final classification."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done 200 out of 200 | elapsed:   18.7s finished\n",
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done 200 out of 200 | elapsed:    2.7s finished\n",
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train score:  0.9994703152073402\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done 200 out of 200 | elapsed:    0.7s finished\n",
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test score: 0.9954214920549421\n",
      "Classification report:                precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      0.99      1.00     19264\n",
      "           1       0.99      1.00      1.00     17866\n",
      "\n",
      "    accuracy                           1.00     37130\n",
      "   macro avg       1.00      1.00      1.00     37130\n",
      "weighted avg       1.00      1.00      1.00     37130\n",
      "\n",
      "Average weighted precision:  0.9961938878316355\n",
      "Average weighted recall:  0.9943016759776536\n",
      "Average weighted f1 score:  0.9952468825141195\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done 200 out of 200 | elapsed:    0.6s finished\n"
     ]
    }
   ],
   "source": [
    "rf_mod = RandomForestClassifier(n_estimators=200, criterion='entropy', random_state= 10, verbose= 1)\n",
    "rf_mod.fit(X_train, y_train)\n",
    "\n",
    "print(\"Train score: \",rf_mod.score(X_train, y_train))\n",
    "\n",
    "pred_rf = rf_mod.predict(X_test)\n",
    "print('Test score:',rf_mod.score(X_test, y_test))\n",
    "\n",
    "print(\"Classification report: \",classification_report(pred_rf,y_test))\n",
    "\n",
    "print(\"Average weighted precision: \",precision_score(y_test,pred_rf))\n",
    "print(\"Average weighted recall: \",recall_score(y_test,pred_rf))\n",
    "print(\"Average weighted f1 score: \",f1_score(y_test,pred_rf))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Support Vector Machine\n",
    "\n",
    "The objective of the support vector machine is to find a Hyper-plane in an N-dimensional feature space, which distinctly classifies the target. The hyper-plane chosen is such that it has maximum margin, i.e. it has maximum distance from all the two classes. This gives enough room for future unseen data to be classified correctly."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train score:  0.9836875039277474\n",
      "Test score: 0.9828709938055481\n",
      "Classification report:                precision    recall  f1-score   support\n",
      "\n",
      "           0       0.99      0.98      0.98     19410\n",
      "           1       0.98      0.99      0.98     17720\n",
      "\n",
      "    accuracy                           0.98     37130\n",
      "   macro avg       0.98      0.98      0.98     37130\n",
      "weighted avg       0.98      0.98      0.98     37130\n",
      "\n",
      "Average weighted precision:  0.9871331828442438\n",
      "Average weighted recall:  0.9772067039106145\n",
      "Average weighted f1 score:  0.9821448624368333\n"
     ]
    }
   ],
   "source": [
    "svm_mod = SVC(probability=True)\n",
    "svm_mod.fit(X_train, y_train)\n",
    "\n",
    "print(\"Train score: \",svm_mod.score(X_train, y_train))\n",
    "\n",
    "pred_svc = svm_mod.predict(X_test)\n",
    "print('Test score:',svm_mod.score(X_test, y_test))\n",
    "\n",
    "print(\"Classification report: \",classification_report(pred_svc,y_test))\n",
    "\n",
    "print(\"Average weighted precision: \",precision_score(y_test,pred_svc))\n",
    "print(\"Average weighted recall: \",recall_score(y_test,pred_svc))\n",
    "print(\"Average weighted f1 score: \",f1_score(y_test,pred_svc))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Logistic Regression\n",
    "\n",
    "Logistic regression applies Linear regression on the data by fitting an equation of similar kind. The predictions are trasnformed by the sigmoid function, which restricts the value between 0 and 1. If the output is a value larger than 0.5, the classifier outputs a 1, otherwise a 0."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train score:  0.9361954267553664\n",
      "Test score: 0.9373552383517372\n",
      "Classification report:                precision    recall  f1-score   support\n",
      "\n",
      "           0       0.96      0.93      0.94     19840\n",
      "           1       0.92      0.95      0.93     17290\n",
      "\n",
      "    accuracy                           0.94     37130\n",
      "   macro avg       0.94      0.94      0.94     37130\n",
      "weighted avg       0.94      0.94      0.94     37130\n",
      "\n",
      "Average weighted precision:  0.9503759398496241\n",
      "Average weighted recall:  0.9179888268156424\n",
      "Average weighted f1 score:  0.9339016766126741\n"
     ]
    }
   ],
   "source": [
    "lg_mod = LogisticRegression()\n",
    "lg_mod.fit(X_train, y_train)\n",
    "\n",
    "print(\"Train score: \",lg_mod.score(X_train, y_train))\n",
    "\n",
    "pred_lg = lg_mod.predict(X_test)\n",
    "print('Test score:',lg_mod.score(X_test, y_test))\n",
    "\n",
    "print(\"Classification report: \",classification_report(pred_lg,y_test))\n",
    "\n",
    "print(\"Average weighted precision: \",precision_score(y_test,pred_lg))\n",
    "print(\"Average weighted recall: \",recall_score(y_test,pred_lg))\n",
    "print(\"Average weighted f1 score: \",f1_score(y_test,pred_lg))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Gaussian Naive Bayes\n",
    "\n",
    "Naive Bayes is a simple algorthim, that can predict the relative probability of a target class and thus classify them. It uses the Bayes theorem to calculate the probability of an input data point belonging to a target class [ P(target class|input) ]. It assumes there exist no correlation between the features and thus is Naive."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train score:  0.8799680393582734\n",
      "Test score: 0.8808241314301104\n",
      "Classification report:                precision    recall  f1-score   support\n",
      "\n",
      "           0       0.89      0.88      0.89     19399\n",
      "           1       0.87      0.88      0.88     17731\n",
      "\n",
      "    accuracy                           0.88     37130\n",
      "   macro avg       0.88      0.88      0.88     37130\n",
      "weighted avg       0.88      0.88      0.88     37130\n",
      "\n",
      "Average weighted precision:  0.87998420844848\n",
      "Average weighted recall:  0.8716759776536313\n",
      "Average weighted f1 score:  0.8758103898290813\n"
     ]
    }
   ],
   "source": [
    "gnb_mod = GaussianNB()\n",
    "gnb_mod.fit(X_train, y_train)\n",
    "\n",
    "print(\"Train score: \",gnb_mod.score(X_train, y_train))\n",
    "\n",
    "pred_nb = gnb_mod.predict(X_test)\n",
    "print('Test score:',gnb_mod.score(X_test, y_test))\n",
    "\n",
    "print(\"Classification report: \",classification_report(pred_nb,y_test))\n",
    "\n",
    "print(\"Average weighted precision: \",precision_score(y_test,pred_nb))\n",
    "print(\"Average weighted recall: \",recall_score(y_test,pred_nb))\n",
    "print(\"Average weighted f1 score: \",f1_score(y_test,pred_nb))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Gradient Boosting\n",
    "\n",
    "The Gradient Boosting classifier, an ensemble technique, is based on the idea of creating several weak learners (A learner which is slightly better than random chance) to minimize loss of the model via a gradient descent like procedure. Every learner learns on a different diffcult pattern and together a majority vote predicts the target class."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train score:  0.9926382791528634\n",
      "Test score: 0.99130083490439\n",
      "Classification report:                precision    recall  f1-score   support\n",
      "\n",
      "           0       0.99      0.99      0.99     19265\n",
      "           1       0.99      0.99      0.99     17865\n",
      "\n",
      "    accuracy                           0.99     37130\n",
      "   macro avg       0.99      0.99      0.99     37130\n",
      "weighted avg       0.99      0.99      0.99     37130\n",
      "\n",
      "Average weighted precision:  0.9919395465994962\n",
      "Average weighted recall:  0.99\n",
      "Average weighted f1 score:  0.9909688242695373\n"
     ]
    }
   ],
   "source": [
    "gb_mod = GradientBoostingClassifier(n_estimators=200, random_state= 10)\n",
    "gb_mod.fit(X_train, y_train)\n",
    "\n",
    "print(\"Train score: \",gb_mod.score(X_train, y_train))\n",
    "\n",
    "pred_gb = gb_mod.predict(X_test)\n",
    "print('Test score:',gb_mod.score(X_test, y_test))\n",
    "\n",
    "print(\"Classification report: \",classification_report(pred_gb,y_test))\n",
    "\n",
    "print(\"Average weighted precision: \",precision_score(y_test,pred_gb))\n",
    "print(\"Average weighted recall: \",recall_score(y_test,pred_gb))\n",
    "print(\"Average weighted f1 score: \",f1_score(y_test,pred_gb))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Multi-Layer Perceptron\n",
    "\n",
    "A Multi-Layer Perceptron classifier is a neural network classifier that applies the Error Back Propogation Training (EBPT) algorithm to learn weights for the network that suitably classify the input pattern space to target classes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train score:  0.9945056424896981\n",
      "Test score: 0.9905467277134392\n",
      "Classification report:                precision    recall  f1-score   support\n",
      "\n",
      "           0       0.99      0.99      0.99     19095\n",
      "           1       0.99      0.99      0.99     18035\n",
      "\n",
      "    accuracy                           0.99     37130\n",
      "   macro avg       0.99      0.99      0.99     37130\n",
      "weighted avg       0.99      0.99      0.99     37130\n",
      "\n",
      "Average weighted precision:  0.9865261990573884\n",
      "Average weighted recall:  0.9939664804469274\n",
      "Average weighted f1 score:  0.9902323639905384\n"
     ]
    }
   ],
   "source": [
    "mlp_mod = MLPClassifier(hidden_layer_sizes=[100,50])\n",
    "mlp_mod.fit(X_train, y_train)\n",
    "\n",
    "print(\"Train score: \",mlp_mod.score(X_train, y_train))\n",
    "\n",
    "pred_mlp = mlp_mod.predict(X_test)\n",
    "print('Test score:',mlp_mod.score(X_test, y_test))\n",
    "\n",
    "print(\"Classification report: \",classification_report(pred_mlp,y_test))\n",
    "\n",
    "print(\"Average weighted precision: \",precision_score(y_test,pred_mlp))\n",
    "print(\"Average weighted recall: \",recall_score(y_test,pred_mlp))\n",
    "print(\"Average weighted f1 score: \",f1_score(y_test,pred_mlp))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model Comparision\n",
    "\n",
    "All the 6 models show good accuracy, precision, recall and f1 scores on the test set. The models are said to have trained well and show a good fit. To evaluate these models, the ROC curve and Calibration plots have been ploted below.<br>\n",
    "\n",
    "The ROC curve, shows the True Positive rates vs the False positive rates for the model. The closer the area under the curve is to a 1, the better the model trained is. The reference is a no skill model, which predicts only one class all the time and has an area under the curve of 0.5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "No Skill AUC:  0.5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done 200 out of 200 | elapsed:    0.9s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RandomForest AUC:  0.999834521537286\n",
      "Support Vector Machine AUC:  0.9973159968856854\n",
      "Logistic Regression AUC:  0.9821765209155853\n",
      "Gaussian Naive Bayes AUC:  0.9420016065447085\n",
      "GradientBoosting AUC:  0.99929816365839\n",
      "MultiLayer Perceptron AUC:  0.999269683077826\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7ff8a0af2b90>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "no_skill_prob = [0 for _ in range(len(y_test))]\n",
    "no_skill_auc = roc_auc_score(y_test, no_skill_prob)\n",
    "print(\"No Skill AUC: \", no_skill_auc)\n",
    "ns_fpr, ns_tpr, _ = roc_curve(y_test, no_skill_prob)\n",
    "\n",
    "rf_prob = rf_mod.predict_proba(X_test)[:,-1]\n",
    "rf_auc = roc_auc_score(y_test,rf_prob)\n",
    "print(\"RandomForest AUC: \", rf_auc)\n",
    "rf_fpr, rf_tpr, _ = roc_curve(y_test, rf_prob)\n",
    "\n",
    "svm_prob = svm_mod.predict_proba(X_test)[:,-1]\n",
    "svm_auc = roc_auc_score(y_test,svm_prob)\n",
    "print(\"Support Vector Machine AUC: \", svm_auc)\n",
    "svm_fpr, svm_tpr, _ = roc_curve(y_test, svm_prob)\n",
    "\n",
    "lg_prob = lg_mod.predict_proba(X_test)[:,-1]\n",
    "lg_auc = roc_auc_score(y_test, lg_prob)\n",
    "print(\"Logistic Regression AUC: \",lg_auc)\n",
    "lg_fpr, lg_tpr, _ = roc_curve(y_test, lg_prob)\n",
    "\n",
    "nb_prob = gnb_mod.predict_proba(X_test)[:,-1]\n",
    "nb_auc = roc_auc_score(y_test, nb_prob)\n",
    "print(\"Gaussian Naive Bayes AUC: \", nb_auc)\n",
    "nb_fpr, nb_tpr, _ = roc_curve(y_test, nb_prob)\n",
    "\n",
    "gb_prob = gb_mod.predict_proba(X_test)[:,-1]\n",
    "gb_auc = roc_auc_score(y_test, gb_prob)\n",
    "print(\"GradientBoosting AUC: \", gb_auc)\n",
    "gb_fpr, gb_tpr, _ = roc_curve(y_test, gb_prob)\n",
    "\n",
    "mlp_prob = mlp_mod.predict_proba(X_test)[:,-1]\n",
    "mlp_auc = roc_auc_score(y_test, mlp_prob)\n",
    "print(\"MultiLayer Perceptron AUC: \", mlp_auc)\n",
    "mlp_fpr, mlp_tpr, _ = roc_curve(y_test, mlp_prob)\n",
    "\n",
    "fig,ax = plt.subplots(figsize=(12,8))\n",
    "ax.plot(ns_fpr, ns_tpr, linestyle= '--', label= 'Reference')\n",
    "ax.plot(rf_fpr, rf_tpr, linestyle= '-', label= 'RandomForest')\n",
    "ax.plot(svm_fpr, svm_tpr, linestyle= '-', label= 'SupportVectorMachine')\n",
    "ax.plot(lg_fpr, lg_tpr, linestyle='-', label= 'LogisticRegression')\n",
    "ax.plot(nb_fpr, nb_tpr, linestyle='-', label= 'GaussianNB')\n",
    "ax.plot(gb_fpr, gb_tpr, linestyle='-', label= 'GradientBoosting')\n",
    "ax.plot(mlp_fpr, mlp_tpr, linestyle='-', label= 'MultiLayer Perceptron')\n",
    "plt.xlabel('False Positve rate')\n",
    "plt.ylabel('True Positive rate')\n",
    "plt.title('ROC curve')\n",
    "ax.legend(loc=\"bottom right\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The below plot is a zoomed in version of the ROC curve plot, to clearly distinguish between the models."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7ff89faa2810>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig,ax = plt.subplots(figsize=(12,8))\n",
    "ax.plot(ns_fpr, ns_tpr, linestyle= '--', label= 'Reference')\n",
    "ax.plot(rf_fpr, rf_tpr, linestyle= '-', label= 'RandomForest')\n",
    "ax.plot(svm_fpr, svm_tpr, linestyle= '-', label= 'SupportVectorMachine')\n",
    "ax.plot(lg_fpr, lg_tpr, linestyle='-', label= 'LogisticRegression')\n",
    "ax.plot(nb_fpr, nb_tpr, linestyle='-', label= 'GaussianNB')\n",
    "ax.plot(gb_fpr, gb_tpr, linestyle='-', label= 'GradientBoosting')\n",
    "ax.plot(mlp_fpr, mlp_tpr, linestyle='-', label= 'MultiLayer Perceptron')\n",
    "plt.xlabel('False Positve rate')\n",
    "plt.ylabel('True Positive rate')\n",
    "plt.title('ROC curve (Zoomed on top left)')\n",
    "plt.xlim([0,0.1])\n",
    "plt.ylim([0.9,1])\n",
    "ax.legend(loc=\"bottom right\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A reliability diagram is a line plot of the relative frequency of what was observed (y-axis) versus the predicted probability frequency (x-axis). The predicted probabilites are divided into fixed number of buckets (bins). The more reliable predictions are, the closer they are to being on the line y=x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "rf_x, rf_y = calibration_curve(y_test, rf_prob, n_bins=10)\n",
    "sv_x, sv_y = calibration_curve(y_test, svm_prob, n_bins=10)\n",
    "lg_x, lg_y = calibration_curve(y_test, lg_prob, n_bins=10)\n",
    "nb_x, nb_y = calibration_curve(y_test, nb_prob, n_bins=10)\n",
    "gb_x, gb_y = calibration_curve(y_test, gb_prob, n_bins=10)\n",
    "mlp_x, mlp_y = calibration_curve(y_test, mlp_prob, n_bins=10)\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(12,8))\n",
    "ax.plot(rf_x, rf_y, marker= 'o', linewidth= '1', label= 'RandomForest')\n",
    "ax.plot(sv_x, sv_y, marker= 'o', linewidth= '1', label= 'SupportVector Machine')\n",
    "# ax.plot(lg_x, lg_y, marker= 'o', linewidth= '1', label= 'LogisticRegression')\n",
    "# ax.plot(nb_x, nb_y, marker= 'o', linewidth= '1', label= 'GaussianNB')\n",
    "ax.plot(gb_x, gb_y, marker= 'o', linewidth= '1', label= 'GradientBoosting')\n",
    "ax.plot(mlp_x, mlp_y, marker= 'o', linewidth= '1', label= 'MultiLayer Perceptron')\n",
    "\n",
    "line = mlines.Line2D([0,1],[0,1], color='black')\n",
    "ax.add_line(line)\n",
    "ax.legend()\n",
    "plt.xlabel('Mean predicted probability')\n",
    "plt.ylabel('Fraction of positives')\n",
    "plt.title('Reliability (Calibration curve)')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The table below tabulates all the scores from each model for a clear comparision."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done 200 out of 200 | elapsed:    0.8s finished\n",
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done 200 out of 200 | elapsed:    0.8s finished\n"
     ]
    }
   ],
   "source": [
    "models = {\n",
    "    'Random Forest':rf_mod, \n",
    "    'Support Vector Machine': svm_mod,\n",
    "    'Logistic Regression':lg_mod,\n",
    "    'Gaussian Naive Bayes': gnb_mod,\n",
    "    'Gradient Boosting':gb_mod,\n",
    "    'Multi-Layer Perceptron':mlp_mod\n",
    "}\n",
    "\n",
    "columns = [\n",
    "    'accuracy',\n",
    "    'precision',\n",
    "    'recall',\n",
    "    'f1'\n",
    "]\n",
    "\n",
    "table = {}\n",
    "\n",
    "\n",
    "for model in models.keys():\n",
    "    \n",
    "    pred = models[model].predict(X_test)\n",
    "    accuracy = models[model].score(X_test,y_test)\n",
    "    precision = precision_score(y_test,pred)\n",
    "    recall = recall_score(y_test, pred)\n",
    "    f1 = f1_score(y_test,pred)\n",
    "    \n",
    "    table[model] = [accuracy, precision, recall, f1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "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>accuracy</th>\n",
       "      <th>precision</th>\n",
       "      <th>recall</th>\n",
       "      <th>f1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Random Forest</th>\n",
       "      <td>0.995421</td>\n",
       "      <td>0.996194</td>\n",
       "      <td>0.994302</td>\n",
       "      <td>0.995247</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Support Vector Machine</th>\n",
       "      <td>0.982871</td>\n",
       "      <td>0.987133</td>\n",
       "      <td>0.977207</td>\n",
       "      <td>0.982145</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Logistic Regression</th>\n",
       "      <td>0.937355</td>\n",
       "      <td>0.950376</td>\n",
       "      <td>0.917989</td>\n",
       "      <td>0.933902</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Gaussian Naive Bayes</th>\n",
       "      <td>0.880824</td>\n",
       "      <td>0.879984</td>\n",
       "      <td>0.871676</td>\n",
       "      <td>0.875810</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Gradient Boosting</th>\n",
       "      <td>0.991301</td>\n",
       "      <td>0.991940</td>\n",
       "      <td>0.990000</td>\n",
       "      <td>0.990969</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Multi-Layer Perceptron</th>\n",
       "      <td>0.990547</td>\n",
       "      <td>0.986526</td>\n",
       "      <td>0.993966</td>\n",
       "      <td>0.990232</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        accuracy  precision    recall        f1\n",
       "Random Forest           0.995421   0.996194  0.994302  0.995247\n",
       "Support Vector Machine  0.982871   0.987133  0.977207  0.982145\n",
       "Logistic Regression     0.937355   0.950376  0.917989  0.933902\n",
       "Gaussian Naive Bayes    0.880824   0.879984  0.871676  0.875810\n",
       "Gradient Boosting       0.991301   0.991940  0.990000  0.990969\n",
       "Multi-Layer Perceptron  0.990547   0.986526  0.993966  0.990232"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results = pd.DataFrame.from_dict(table, orient='index')\n",
    "results.columns = columns\n",
    "results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Inference\n",
    "\n",
    "After training all the six models, listed below:- \n",
    "\n",
    "* Random Forest\n",
    "* Support Vector Machine\n",
    "* Logistic Regression\n",
    "* Gaussian Naive Bayes\n",
    "* Gradient Boosting\n",
    "* Multi-Layer Perceptron\n",
    "\n",
    "From these trained models, the results table shows the Random Forest, Gradient Boosting and Multi-Layer Perceptron classifiers boast a 99% accuracy, precision, recall and f1 scores. The Support Vector machine is not far behind with a 98%.<br>\n",
    "The ROC curve shows, similar results that these three models - Random Forest, Gradient Boosting and Multi-Layer Perceptron have about 0.99 area under the curve. The Random Forest among these three has the highest Area under the curve.<br>\n",
    "The Reliability diagram (Calibration curve) shows that Random Forest and Gradient Boosting classifiers mimic the diagonal (y=x) line the most and hence tend to output reliable predictions. The Multi-Perceptron classifier until now showed at par results, but the calibration curve shows that its way off as compared to the other two and hence the predictions from this model are not reliable.<br>\n",
    "\n",
    "In conclusion, the Random Forest classifier out performs all the models in detecting intrusion on the network. Gradient Boosting classifiers are a close second in detecting intrusions and giving reliable predictions.\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}