{ "cells": [ { "cell_type": "markdown", "metadata": { "tags": [] }, "source": [ "> **Essential ML process for Intrusion Detection**\n", "
` python 3.7.13 scikit-learn 1.0.2 `\n", "
`numpy 1.19.5 pandas 1.3.5`\n", "
DETERMINE THE BEST ALGORITHM FOR INTRUSION DETECTION " ] }, { "cell_type": "markdown", "metadata": { "colab_type": "text", "id": "_S34U5S-i69d" }, "source": [ "**Import the main libraries**" ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "import numpy\n", "import pandas\n", "\n", "from time import time\n", "\n", "import os\n", "data_path = '../datasets/NSL_KDD'" ] }, { "cell_type": "markdown", "metadata": { "tags": [] }, "source": [ "_import the local library_" ] }, { "cell_type": "code", "execution_count": 2, "metadata": { "tags": [] }, "outputs": [], "source": [ "# add parent folder path where lib folder is\n", "import sys\n", "if \"..\" not in sys.path:import sys; sys.path.insert(0, '..') " ] }, { "cell_type": "code", "execution_count": 3, "metadata": { "tags": [] }, "outputs": [], "source": [ "from mylib import show_labels_dist, show_metrics, bias_var_metrics" ] }, { "cell_type": "markdown", "metadata": { "colab_type": "text", "id": "9IZetEZ8jQJm" }, "source": [ "**Import the Dataset**\n", "Explore the size of dataset" ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Train Dataset: 63280 rows, 43 columns\n", "Test Dataset: 22544 rows, 43 columns\n" ] } ], "source": [ "# Using boosted Train and preprocessed Test\n", "\n", "data_file = os.path.join(data_path, 'NSL_boosted-2.csv') \n", "train_df = pandas.read_csv(data_file)\n", "print('Train Dataset: {} rows, {} columns'.format(train_df.shape[0], train_df.shape[1]))\n", "\n", "data_file = os.path.join(data_path, 'NSL_ppTest.csv') \n", "test_df = pandas.read_csv(data_file)\n", "print('Test Dataset: {} rows, {} columns'.format(test_df.shape[0], test_df.shape[1]))" ] }, { "cell_type": "markdown", "metadata": { "colab_type": "text", "id": "Cpzuyj7gxwCg" }, "source": [ "***\n", "**Data Preparation and Exploratory Data Analysis (EDA)** " ] }, { "cell_type": "markdown", "metadata": { "tags": [] }, "source": [ "* _Check column names of numeric attributes_" ] }, { "cell_type": "code", "execution_count": 5, "metadata": { "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Numeric features in the train_set that are not in the test_set: None\n", "Numeric features in the test_set that are not in the train_set: None\n", "\n" ] } ], "source": [ "# Ensure both train and test datasets have consistent numeric features\n", "trnn = train_df.select_dtypes(include=['float64','int64']).columns\n", "tstn = test_df.select_dtypes(include=['float64','int64']).columns\n", "trndif = numpy.setdiff1d(trnn, tstn)\n", "tstdif = numpy.setdiff1d(tstn, trnn)\n", "\n", "print(\"Numeric features in the train_set that are not in the test_set: \",end='')\n", "if len(trndif) > 0:\n", " print('\\n',trndif)\n", "else:\n", " print('None')\n", "\n", "print(\"Numeric features in the test_set that are not in the train_set: \",end='')\n", "if len(tstdif) > 0:\n", " print('\\n',tstdif)\n", "else:\n", " print('None')\n", "\n", "print()\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "* _Check column names of categorical attributes_" ] }, { "cell_type": "code", "execution_count": 6, "metadata": { "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Categorical features in the train_set that are not in the test_set: None\n", "Categorical features in the test_set that are not in the train_set: None\n", "\n" ] } ], "source": [ "# Ensure both train and test datasets have consistent categorical features\n", "trnn = train_df.select_dtypes(include=['object']).columns\n", "tstn = test_df.select_dtypes(include=['object']).columns\n", "trndif = numpy.setdiff1d(trnn, tstn)\n", "tstdif = numpy.setdiff1d(tstn, trnn)\n", "\n", "print(\"Categorical features in the train_set that are not in the test_set: \",end='')\n", "if len(trndif) > 0:\n", " print('\\n',trndif)\n", "else:\n", " print('None')\n", "\n", "print(\"Categorical features in the test_set that are not in the train_set: \",end='')\n", "if len(tstdif) > 0:\n", " print('\\n\\t',tstdif)\n", "else:\n", " print('None')\n", "\n", "print()\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "* _Check for missing values_" ] }, { "cell_type": "code", "execution_count": 7, "metadata": { "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Missing Values - Train Set\n", "Total 0 features with null values\n", "Missing Values - Test Set\n", "Total 0 features with null values\n" ] } ], "source": [ "cnt=0\n", "print('Missing Values - Train Set')\n", "for col in train_df.columns:\n", "# print(col, ' ::> ', len(combined_df[col].unique()))\n", " nnul = pandas.notnull(train_df[col]) \n", " if (len(nnul)!=len(train_df)):\n", " cnt=cnt+1\n", " print('\\t',col,':',(len(test_df)-len(nnul)),'null values')\n", "print('Total',cnt,'features with null values')\n", "\n", "cnt=0\n", "print('Missing Values - Test Set')\n", "for col in test_df.columns:\n", "# print(col, ' ::> ', len(combined_df[col].unique()))\n", " nnul = pandas.notnull(test_df[col]) \n", " if (len(nnul)!=len(test_df)):\n", " cnt=cnt+1\n", " print('\\t',col,':',(len(test_df)-len(nnul)),'null values')\n", "print('Total',cnt,'features with null values')\n", "\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "* _Quick visual check of unique values, deal with unique identifiers_" ] }, { "cell_type": "code", "execution_count": 8, "metadata": { "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Unique value count: Train ( 63280 rows ) ~ Test( 22544 rows )\n", "duration ::> 1657 ~ 624\n", "protocol_type ::> 3 ~ 3\n", "service ::> 70 ~ 64\n", "flag ::> 11 ~ 11\n", "src_bytes ::> 2553 ~ 1149\n", "dst_bytes ::> 6633 ~ 3650\n", "land ::> 2 ~ 2\n", "wrong_fragment ::> 3 ~ 3\n", "urgent ::> 3 ~ 4\n", "hot ::> 24 ~ 16\n", "num_failed_logins ::> 6 ~ 5\n", "logged_in ::> 2 ~ 2\n", "num_compromised ::> 49 ~ 23\n", "root_shell ::> 2 ~ 2\n", "su_attempted ::> 3 ~ 3\n", "num_root ::> 44 ~ 20\n", "num_file_creations ::> 28 ~ 9\n", "num_shells ::> 3 ~ 4\n", "num_access_files ::> 9 ~ 5\n", "num_outbound_cmds ::> 1 ~ 1\n", "is_host_login ::> 2 ~ 2\n", "is_guest_login ::> 2 ~ 2\n", "count ::> 511 ~ 495\n", "srv_count ::> 489 ~ 457\n", "serror_rate ::> 87 ~ 88\n", "srv_serror_rate ::> 75 ~ 82\n", "rerror_rate ::> 81 ~ 90\n", "srv_rerror_rate ::> 63 ~ 93\n", "same_srv_rate ::> 101 ~ 75\n", "diff_srv_rate ::> 88 ~ 99\n", "srv_diff_host_rate ::> 61 ~ 84\n", "dst_host_count ::> 256 ~ 256\n", "dst_host_srv_count ::> 256 ~ 256\n", "dst_host_same_srv_rate ::> 101 ~ 101\n", "dst_host_diff_srv_rate ::> 101 ~ 101\n", "dst_host_same_src_port_rate ::> 101 ~ 101\n", "dst_host_srv_diff_host_rate ::> 67 ~ 58\n", "dst_host_serror_rate ::> 101 ~ 99\n", "dst_host_srv_serror_rate ::> 100 ~ 101\n", "dst_host_rerror_rate ::> 101 ~ 101\n", "dst_host_srv_rerror_rate ::> 101 ~ 100\n", "label ::> 40 ~ 38\n", "atakcat ::> 5 ~ 5\n" ] } ], "source": [ "# Identify columns with only one value \n", "# or with number of unique values == number of rows\n", "n_eq_one = []\n", "n_eq_all = []\n", "\n", "print('Unique value count: Train (',train_df.shape[0],'rows ) ~ Test(',test_df.shape[0],'rows )')\n", "for col in train_df.columns:\n", " lctrn = len(train_df[col].unique())\n", " lctst = len(test_df[col].unique())\n", " print(col, ' ::> ', lctrn, ' ~ ', lctst)\n", " if (lctrn == 1) and (lctrn == lctst): \n", " n_eq_one.append(train_df[col].name)\n", " if lctrn == train_df.shape[0]:\n", " n_eq_all.append(train_df[col].name)" ] }, { "cell_type": "code", "execution_count": 9, "metadata": { "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Dropping single-valued features\n", "['num_outbound_cmds']\n" ] } ], "source": [ "# Drop columns with only one value\n", "# Drop unique identifier\n", "if len(n_eq_one) > 0:\n", " print('Dropping single-valued features')\n", " print(n_eq_one)\n", " train_df.drop(n_eq_one, axis=1, inplace=True)\n", " test_df.drop(n_eq_one, axis=1, inplace=True)\n", "\n", "# Drop or bin columns with number of unique values == number of rows\n", "if len(n_eq_all) > 0:\n", " print('Dropping unique identifiers')\n", " print(n_eq_all)\n", " train_df.drop(n_eq_all, axis=1, inplace=True)\n", " test_df.drop(n_eq_all, axis=1, inplace=True)\n", "\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "* _Check categorical feature values:
\n", "differences will be resolved by one-hot encoding the combined test and train sets_" ] }, { "cell_type": "code", "execution_count": 10, "metadata": { "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "protocol_type ::> \n", "\tUnique text values in the train_set that are not in the test_set: None\n", "\tUnique text values in the test_set that are not in the train_set: None\n", "service ::> \n", "\tUnique text values in the train_set that are not in the test_set: \n", "\t ['aol' 'harvest' 'http_2784' 'http_8001' 'red_i' 'urh_i']\n", "\tUnique text values in the test_set that are not in the train_set: None\n", "flag ::> \n", "\tUnique text values in the train_set that are not in the test_set: None\n", "\tUnique text values in the test_set that are not in the train_set: None\n", "label ::> \n", "\tUnique text values in the train_set that are not in the test_set: \n", "\t ['spy' 'warezclient']\n", "\tUnique text values in the test_set that are not in the train_set: None\n", "atakcat ::> \n", "\tUnique text values in the train_set that are not in the test_set: None\n", "\tUnique text values in the test_set that are not in the train_set: None\n" ] } ], "source": [ "# identify any inconsistencies between the train and test datasets \n", "# because it could potentially affect model training and evaluation.\n", "trnn = train_df.select_dtypes(include=['object']).columns\n", "for col in trnn:\n", " tr = train_df[col].unique()\n", " ts = test_df[col].unique()\n", " trd = numpy.setdiff1d(tr, ts)\n", " tsd = numpy.setdiff1d(ts, tr)\n", " \n", " print(col,'::> ')\n", " print(\"\\tUnique text values in the train_set that are not in the test_set: \",end='')\n", " if len(trd) > 0:\n", " print('\\n\\t',trd)\n", " else:\n", " print('None')\n", " \n", " print(\"\\tUnique text values in the test_set that are not in the train_set: \",end='')\n", " if len(tsd) > 0:\n", " print('\\n\\t',tsd)\n", " else:\n", " print('None')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "* _Combine for processing classification target and text features_" ] }, { "cell_type": "code", "execution_count": 11, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Combined Dataset: 85824 rows, 42 columns\n" ] } ], "source": [ "# Explore the size of train and test dataset combined\n", "combined_df = pandas.concat([train_df, test_df])\n", "print('Combined Dataset: {} rows, {} columns'.format(\n", " combined_df.shape[0], combined_df.shape[1]))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "* _Classification Target feature:_\n", "two columns of labels are available \n", " * Two-class: Reduce the detailed attack labels to 'normal' or 'attack'\n", " * Multiclass: Use the category labels (atakcat)" ] }, { "cell_type": "code", "execution_count": 12, "metadata": { "tags": [] }, "outputs": [ { "data": { "text/plain": [ "label\n", "normal 43383\n", "neptune 25264\n", "satan 2552\n", "smurf 1988\n", "ipsweep 1941\n", "portsweep 1623\n", "guess_passwd 1257\n", "mscan 1046\n", "warezmaster 954\n", "back 837\n", "nmap 820\n", "apache2 774\n", "processtable 719\n", "teardrop 458\n", "warezclient 445\n", "snmpguess 364\n", "saint 350\n", "mailbomb 322\n", "snmpgetattack 196\n", "httptunnel 146\n", "pod 142\n", "buffer_overflow 35\n", "named 25\n", "ps 22\n", "multihop 21\n", "sendmail 21\n", "xterm 20\n", "rootkit 18\n", "land 16\n", "xlock 14\n", "xsnoop 8\n", "ftp_write 7\n", "phf 6\n", "loadmodule 6\n", "imap 6\n", "perl 5\n", "sqlattack 4\n", "worm 4\n", "udpstorm 4\n", "spy 1\n", "Name: count, dtype: int64" ] }, "execution_count": 12, "metadata": {}, "output_type": "execute_result" } ], "source": [ "combined_df['label'].value_counts()" ] }, { "cell_type": "code", "execution_count": 13, "metadata": { "tags": [] }, "outputs": [ { "data": { "text/plain": [ "atakcat\n", "benign 43383\n", "dos 30524\n", "probe 8332\n", "r2l 3329\n", "u2r 256\n", "Name: count, dtype: int64" ] }, "execution_count": 13, "metadata": {}, "output_type": "execute_result" } ], "source": [ "combined_df['atakcat'].value_counts()" ] }, { "cell_type": "code", "execution_count": 14, "metadata": {}, "outputs": [], "source": [ "# Set the classification target\n", "twoclass = False # 'False' because output has more than two class " ] }, { "cell_type": "code", "execution_count": 15, "metadata": { "tags": [] }, "outputs": [], "source": [ "if twoclass:\n", "# Two-class: Reduce the detailed attack labels to 'normal' or 'attack'\n", "\n", " labels_df = combined_df['label'].copy()\n", " labels_df[labels_df != 'normal'] = 'attack'\n", "else:\n", "# Multiclass: Use the category labels (atakcat)\n", "\n", " labels_df = combined_df[['atakcat']].copy()\n", " labels_df.rename(columns={'atakcat':'label'}, inplace=True)\n", " labels_df = labels_df.squeeze('columns')\n", "\n", "# drop target features \n", "combined_df.drop(['label'], axis=1, inplace=True)\n", "combined_df.drop(['atakcat'], axis=1, inplace=True)" ] }, { "cell_type": "markdown", "metadata": { "tags": [] }, "source": [ "* _One-Hot Encoding the categorical (text) features_" ] }, { "cell_type": "code", "execution_count": 16, "metadata": { "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "['protocol_type', 'service', 'flag']\n" ] } ], "source": [ "# put the names into a python list - for pandas.get_dummies()\n", "categori = combined_df.select_dtypes(include=['object']).columns\n", "category_cols = categori.tolist()\n", "print(category_cols)" ] }, { "cell_type": "code", "execution_count": 17, "metadata": { "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "protocol_type ::> ['icmp' 'tcp' 'udp']\n", "\n", "service ::> ['IRC' 'X11' 'Z39_50' 'aol' 'auth' 'bgp' 'courier' 'csnet_ns' 'ctf'\n", " 'daytime' 'discard' 'domain' 'domain_u' 'echo' 'eco_i' 'ecr_i' 'efs'\n", " 'exec' 'finger' 'ftp' 'ftp_data' 'gopher' 'harvest' 'hostnames' 'http'\n", " 'http_2784' 'http_443' 'http_8001' 'imap4' 'iso_tsap' 'klogin' 'kshell'\n", " 'ldap' 'link' 'login' 'mtp' 'name' 'netbios_dgm' 'netbios_ns'\n", " 'netbios_ssn' 'netstat' 'nnsp' 'nntp' 'ntp_u' 'other' 'pm_dump' 'pop_2'\n", " 'pop_3' 'printer' 'private' 'red_i' 'remote_job' 'rje' 'shell' 'smtp'\n", " 'sql_net' 'ssh' 'sunrpc' 'supdup' 'systat' 'telnet' 'tftp_u' 'tim_i'\n", " 'time' 'urh_i' 'urp_i' 'uucp' 'uucp_path' 'vmnet' 'whois']\n", "\n", "flag ::> ['OTH' 'REJ' 'RSTO' 'RSTOS0' 'RSTR' 'S0' 'S1' 'S2' 'S3' 'SF' 'SH']\n", "\n" ] } ], "source": [ "# generate a sorted list of unique values of categorical features\n", "# we will get a new column for each one with get_dummies()\n", "\n", "for col in categori:\n", " ul = numpy.sort(list(combined_df[col].unique()))\n", " print(col, ' ::> ', ul)\n", " print()" ] }, { "cell_type": "code", "execution_count": 18, "metadata": { "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "Index: 85824 entries, 0 to 22543\n", "Columns: 121 entries, duration to flag_SH\n", "dtypes: bool(84), float64(15), int64(22)\n", "memory usage: 31.8 MB\n" ] } ], "source": [ "# Apply to the list of Categorical columns (text fields)\n", "# one-hot encoding\n", "features_df = pandas.get_dummies(combined_df, columns=category_cols)\n", "features_df.info()" ] }, { "cell_type": "code", "execution_count": 19, "metadata": { "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "['duration', 'src_bytes', 'dst_bytes', 'land', 'wrong_fragment', 'urgent', 'hot', 'num_failed_logins', 'logged_in', 'num_compromised', 'root_shell', 'su_attempted', 'num_root', 'num_file_creations', 'num_shells', 'num_access_files', 'is_host_login', 'is_guest_login', 'count', 'srv_count', 'serror_rate', 'srv_serror_rate', 'rerror_rate', 'srv_rerror_rate', 'same_srv_rate', 'diff_srv_rate', 'srv_diff_host_rate', 'dst_host_count', 'dst_host_srv_count', 'dst_host_same_srv_rate', 'dst_host_diff_srv_rate', 'dst_host_same_src_port_rate', 'dst_host_srv_diff_host_rate', 'dst_host_serror_rate', 'dst_host_srv_serror_rate', 'dst_host_rerror_rate', 'dst_host_srv_rerror_rate']\n" ] } ], "source": [ "# generate a list of numeric columns for scaling - After test // train split\n", "numeri = combined_df.select_dtypes(include=['float64','int64']).columns\n", "print(numeri.to_list())" ] }, { "cell_type": "markdown", "metadata": { "colab_type": "text", "id": "Cpzuyj7gxwCg", "tags": [] }, "source": [ "***\n", "**
Create Test // Train Datasets**\n" ] }, { "cell_type": "code", "execution_count": 20, "metadata": { "tags": [] }, "outputs": [], "source": [ "# Restore the train // test split: slice 1 Dataframe into 2 \n", "features_train = features_df.iloc[:len(train_df),:].copy() # X_train\n", "features_train.reset_index(inplace=True, drop=True)\n", "\n", "# Split features into training and testing dataset\n", "features_test = features_df.iloc[len(train_df):,:].copy() # X_test\n", "features_test.reset_index(inplace=True, drop=True)\n", "\n", "# Split labels into training and testing dataset\n", "labels_train = labels_df[:len(train_df)] # y_train\n", "labels_train.reset_index(inplace=True, drop=True)\n", "\n", "labels_test = labels_df[len(train_df):] # y_test\n", "labels_test.reset_index(inplace=True, drop=True)" ] }, { "cell_type": "markdown", "metadata": { "tags": [] }, "source": [ "***\n", "Next are standard steps for all datasets: _scaling, classifiers, results_" ] }, { "cell_type": "markdown", "metadata": { "colab_type": "text", "id": "Cpzuyj7gxwCg", "tags": [] }, "source": [ "**Scaling** comes _after_ test // train split" ] }, { "cell_type": "code", "execution_count": 21, "metadata": { "tags": [] }, "outputs": [], "source": [ "# scaling the Numeric columns \n", "from sklearn.preprocessing import MinMaxScaler\n", "\n", "\n", "# scaling because certain classifiers work best only with normalized data\n", "# scaling brings feature to similar range\n", "\n", "for i in numeri:\n", " arr = numpy.array(features_train[i])\n", " scale = MinMaxScaler().fit(arr.reshape(-1, 1))\n", " features_train[i] = scale.transform(arr.reshape(len(arr),1))\n", "\n", " arr = numpy.array(features_test[i])\n", " features_test[i] = scale.transform(arr.reshape(len(arr),1))" ] }, { "cell_type": "markdown", "metadata": { "colab_type": "text", "id": "Cpzuyj7gxwCg", "tags": [] }, "source": [ "**
Classifier Selection**" ] }, { "cell_type": "code", "execution_count": 22, "metadata": { "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[('RandomForest', RandomForestClassifier()), ('LogReg', LogisticRegression()), ('SupportVectorClf', SVC()), ('GaussianNB', GaussianNB()), ('K-NNeighbors', KNeighborsClassifier())]\n" ] } ], "source": [ "# prepare list\n", "models = []\n", "\n", "from sklearn.ensemble import RandomForestClassifier \n", "models.append((\"RandomForest\", RandomForestClassifier())) \n", "\n", "from sklearn.linear_model import LogisticRegression \n", "models.append ((\"LogReg\",LogisticRegression())) \n", "\n", "from sklearn.svm import SVC \n", "models.append((\"SupportVectorClf\", SVC())) \n", "\n", "from sklearn.naive_bayes import GaussianNB \n", "models.append ((\"GaussianNB\",GaussianNB())) \n", "\n", "from sklearn.neighbors import KNeighborsClassifier \n", "models.append((\"K-NNeighbors\", KNeighborsClassifier())) \n", "\n", "\n", "\n", "print(models)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "
_compatibility block for pasting in from sample code_" ] }, { "cell_type": "code", "execution_count": 23, "metadata": { "tags": [] }, "outputs": [], "source": [ "# dataset names\n", "X_train = features_train\n", "y_train = labels_train\n", "X_test = features_test\n", "y_test = labels_test\n", "labels_col = 'label'\n", "# library names\n", "pd = pandas\n", "np = numpy" ] }, { "cell_type": "markdown", "metadata": { "tags": [] }, "source": [ "**
Target Label Distributions** (standard block)" ] }, { "cell_type": "code", "execution_count": 24, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "features_train: 63280 rows, 121 columns\n", "features_test: 22544 rows, 121 columns\n", "\n", "labels_train: 63280 rows, 1 column\n", "labels_test: 22544 rows, 1 column\n", "\n", "Frequency and Distribution of labels\n", " label %_train label %_test\n", "label \n", "benign 33672 53.21 9711 43.08\n", "dos 23066 36.45 7458 33.08\n", "probe 5911 9.34 2421 10.74\n", "r2l 575 0.91 2754 12.22\n", "u2r 56 0.09 200 0.89\n" ] } ], "source": [ "# from our local library\n", "# show number of columns and rows in train and test set\n", "show_labels_dist(X_train,X_test,y_train,y_test)" ] }, { "cell_type": "markdown", "metadata": { "colab_type": "text", "id": "Cpzuyj7gxwCg", "tags": [] }, "source": [ "_... this is a good place for the ClassBalance visualizer ..._" ] }, { "cell_type": "markdown", "metadata": { "colab_type": "text", "id": "Cpzuyj7gxwCg" }, "source": [ "**
Fit and Predict** (standard block)" ] }, { "cell_type": "code", "execution_count": 25, "metadata": { "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "macro average: unweighted mean per label\n", "weighted average: support-weighted mean per label\n", "MCC: correlation between prediction and ground truth\n", " (+1 perfect, 0 random prediction, -1 inverse)\n", "\n", "Confusion Matrix: RandomForest\n", "Run Time 14.33 seconds\n", "\n", " pred:benign pred:dos pred:probe pred:r2l pred:u2r\n", "train:benign 9303 67 187 153 1\n", "train:dos 19 7327 112 0 0\n", "train:probe 118 19 2284 0 0\n", "train:r2l 2214 0 1 539 0\n", "train:u2r 59 0 0 0 141\n", "\n", "~~~~\n", " benign : FPR = 0.188 FNR = 0.042\n", " dos : FPR = 0.006 FNR = 0.018\n", " probe : FPR = 0.015 FNR = 0.057\n", " r2l : FPR = 0.008 FNR = 0.804\n", " u2r : FPR = 0.000 FNR = 0.295\n", "\n", " macro avg : FPR = 0.043 FNR = 0.243\n", "weighted avg : FPR = 0.033 FNR = 0.131\n", "\n", "~~~~\n", " precision recall f1-score support\n", "\n", " benign 0.794 0.958 0.868 9711\n", " dos 0.988 0.982 0.985 7458\n", " probe 0.884 0.943 0.913 2421\n", " r2l 0.779 0.196 0.313 2754\n", " u2r 0.993 0.705 0.825 200\n", "\n", " accuracy 0.869 22544\n", " macro avg 0.888 0.757 0.781 22544\n", "weighted avg 0.868 0.869 0.844 22544\n", "\n", "~~~~\n", "MCC: Overall : 0.810\n", " benign : 0.763\n", " dos : 0.978\n", " probe : 0.902\n", " r2l : 0.357\n", " u2r : 0.836\n", "\n", "Parameters: {'bootstrap': True, 'ccp_alpha': 0.0, 'class_weight': None, 'criterion': 'gini', 'max_depth': None, 'max_features': 'sqrt', 'max_leaf_nodes': None, 'max_samples': None, 'min_impurity_decrease': 0.0, 'min_samples_leaf': 1, 'min_samples_split': 2, 'min_weight_fraction_leaf': 0.0, 'n_estimators': 100, 'n_jobs': None, 'oob_score': False, 'random_state': None, 'verbose': 0, 'warm_start': False} \n", "\n", "\n", "Confusion Matrix: LogReg\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Run Time 5.51 seconds\n", "\n", " pred:benign pred:dos pred:probe pred:r2l pred:u2r\n", "train:benign 9004 96 608 1 2\n", "train:dos 1176 6257 25 0 0\n", "train:probe 257 85 2073 6 0\n", "train:r2l 2664 4 2 80 4\n", "train:u2r 177 1 2 4 16\n", "\n", "~~~~\n", " benign : FPR = 0.333 FNR = 0.073\n", " dos : FPR = 0.012 FNR = 0.161\n", " probe : FPR = 0.032 FNR = 0.144\n", " r2l : FPR = 0.001 FNR = 0.971\n", " u2r : FPR = 0.000 FNR = 0.920\n", "\n", " macro avg : FPR = 0.076 FNR = 0.454\n", "weighted avg : FPR = 0.057 FNR = 0.227\n", "\n", "~~~~\n", " precision recall f1-score support\n", "\n", " benign 0.678 0.927 0.783 9711\n", " dos 0.971 0.839 0.900 7458\n", " probe 0.765 0.856 0.808 2421\n", " r2l 0.879 0.029 0.056 2754\n", " u2r 0.727 0.080 0.144 200\n", "\n", " accuracy 0.773 22544\n", " macro avg 0.804 0.546 0.538 22544\n", "weighted avg 0.809 0.773 0.730 22544\n", "\n", "~~~~\n", "MCC: Overall : 0.669\n", " benign : 0.598\n", " dos : 0.861\n", " probe : 0.785\n", " r2l : 0.147\n", " u2r : 0.239\n", "\n", "Parameters: {'C': 1.0, 'class_weight': None, 'dual': False, 'fit_intercept': True, 'intercept_scaling': 1, 'l1_ratio': None, 'max_iter': 100, 'multi_class': 'auto', 'n_jobs': None, 'penalty': 'l2', 'random_state': None, 'solver': 'lbfgs', 'tol': 0.0001, 'verbose': 0, 'warm_start': False} \n", "\n", "\n", "Confusion Matrix: SupportVectorClf\n", "Run Time 90.74 seconds\n", "\n", " pred:benign pred:dos pred:probe pred:r2l pred:u2r\n", "train:benign 9015 59 630 0 7\n", "train:dos 643 6759 56 0 0\n", "train:probe 182 144 2091 0 4\n", "train:r2l 2407 216 8 119 4\n", "train:u2r 60 0 0 3 137\n", "\n", "~~~~\n", " benign : FPR = 0.257 FNR = 0.072\n", " dos : FPR = 0.028 FNR = 0.094\n", " probe : FPR = 0.034 FNR = 0.136\n", " r2l : FPR = 0.000 FNR = 0.957\n", " u2r : FPR = 0.001 FNR = 0.315\n", "\n", " macro avg : FPR = 0.064 FNR = 0.315\n", "weighted avg : FPR = 0.049 FNR = 0.196\n", "\n", "~~~~\n", " precision recall f1-score support\n", "\n", " benign 0.733 0.928 0.819 9711\n", " dos 0.942 0.906 0.924 7458\n", " probe 0.751 0.864 0.803 2421\n", " r2l 0.975 0.043 0.083 2754\n", " u2r 0.901 0.685 0.778 200\n", "\n", " accuracy 0.804 22544\n", " macro avg 0.860 0.685 0.681 22544\n", "weighted avg 0.835 0.804 0.762 22544\n", "\n", "~~~~\n", "MCC: Overall : 0.713\n", " benign : 0.668\n", " dos : 0.887\n", " probe : 0.780\n", " r2l : 0.192\n", " u2r : 0.784\n", "\n", "Parameters: {'C': 1.0, 'break_ties': False, 'cache_size': 200, 'class_weight': None, 'coef0': 0.0, 'decision_function_shape': 'ovr', 'degree': 3, 'gamma': 'scale', 'kernel': 'rbf', 'max_iter': -1, 'probability': False, 'random_state': None, 'shrinking': True, 'tol': 0.001, 'verbose': False} \n", "\n", "\n", "Confusion Matrix: GaussianNB\n", "Run Time 1.05 seconds\n", "\n", " pred:benign pred:dos pred:probe pred:r2l pred:u2r\n", "train:benign 4792 58 9 4544 308\n", "train:dos 3138 3017 14 1240 49\n", "train:probe 1193 255 145 516 312\n", "train:r2l 770 0 0 1363 621\n", "train:u2r 3 0 0 10 187\n", "\n", "~~~~\n", " benign : FPR = 0.398 FNR = 0.507\n", " dos : FPR = 0.021 FNR = 0.595\n", " probe : FPR = 0.001 FNR = 0.940\n", " r2l : FPR = 0.319 FNR = 0.505\n", " u2r : FPR = 0.058 FNR = 0.065\n", "\n", " macro avg : FPR = 0.159 FNR = 0.522\n", "weighted avg : FPR = 0.145 FNR = 0.578\n", "\n", "~~~~\n", " precision recall f1-score support\n", "\n", " benign 0.484 0.493 0.489 9711\n", " dos 0.906 0.405 0.559 7458\n", " probe 0.863 0.060 0.112 2421\n", " r2l 0.178 0.495 0.261 2754\n", " u2r 0.127 0.935 0.223 200\n", "\n", " accuracy 0.422 22544\n", " macro avg 0.512 0.478 0.329 22544\n", "weighted avg 0.624 0.422 0.442 22544\n", "\n", "~~~~\n", "MCC: Overall : 0.209\n", " benign : 0.096\n", " dos : 0.509\n", " probe : 0.211\n", " r2l : 0.122\n", " u2r : 0.332\n", "\n", "Parameters: {'priors': None, 'var_smoothing': 1e-09} \n", "\n", "\n", "Confusion Matrix: K-NNeighbors\n", "Run Time 15.25 seconds\n", "\n", " pred:benign pred:dos pred:probe pred:r2l pred:u2r\n", "train:benign 8883 66 620 137 5\n", "train:dos 328 7098 32 0 0\n", "train:probe 217 122 2076 1 5\n", "train:r2l 2039 227 5 477 6\n", "train:u2r 43 0 0 2 155\n", "\n", "~~~~\n", " benign : FPR = 0.205 FNR = 0.085\n", " dos : FPR = 0.028 FNR = 0.048\n", " probe : FPR = 0.033 FNR = 0.143\n", " r2l : FPR = 0.007 FNR = 0.827\n", " u2r : FPR = 0.001 FNR = 0.225\n", "\n", " macro avg : FPR = 0.055 FNR = 0.266\n", "weighted avg : FPR = 0.043 FNR = 0.171\n", "\n", "~~~~\n", " precision recall f1-score support\n", "\n", " benign 0.772 0.915 0.837 9711\n", " dos 0.945 0.952 0.948 7458\n", " probe 0.760 0.857 0.806 2421\n", " r2l 0.773 0.173 0.283 2754\n", " u2r 0.906 0.775 0.836 200\n", "\n", " accuracy 0.829 22544\n", " macro avg 0.831 0.734 0.742 22544\n", "weighted avg 0.829 0.829 0.803 22544\n", "\n", "~~~~\n", "MCC: Overall : 0.748\n", " benign : 0.703\n", " dos : 0.923\n", " probe : 0.782\n", " r2l : 0.333\n", " u2r : 0.837\n", "\n", "Parameters: {'algorithm': 'auto', 'leaf_size': 30, 'metric': 'minkowski', 'metric_params': None, 'n_jobs': None, 'n_neighbors': 5, 'p': 2, 'weights': 'uniform'} \n", "\n", "\n" ] } ], "source": [ "# evaluate each model in turn\n", "results = []\n", "\n", "print('macro average: unweighted mean per label')\n", "print('weighted average: support-weighted mean per label')\n", "print('MCC: correlation between prediction and ground truth')\n", "print(' (+1 perfect, 0 random prediction, -1 inverse)\\n')\n", "\n", "for name, clf in models:\n", " trs = time()\n", " print('Confusion Matrix:', name)\n", " \n", " clf.fit(X_train, y_train)\n", " ygx = clf.predict(X_test)\n", " results.append((name, ygx))\n", " \n", " tre = time() - trs\n", " print (\"Run Time {} seconds\".format(round(tre,2)) + '\\n')\n", " \n", "# the best performing algorithm will be used in proposed framework \n", "\n", " show_metrics(y_test, ygx, clf.classes_) # from our local library\n", " print('\\nParameters: ', clf.get_params(), '\\n\\n')" ] }, { "cell_type": "markdown", "metadata": { "colab_type": "text", "id": "Cpzuyj7gxwCg", "tags": [] }, "source": [ "**Bias - Variance Decomposition** (standard block)" ] }, { "cell_type": "code", "execution_count": 26, "metadata": { "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Bias // Variance Decomposition: RandomForest\n", " Average bias: 0.132\n", " Average variance: 0.016\n", " Average expected loss: 0.134 \"Goodness\": 0.866\n", "\n", "Bias // Variance Decomposition: LogReg\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ " Average bias: 0.227\n", " Average variance: 0.008\n", " Average expected loss: 0.226 \"Goodness\": 0.774\n", "\n", "Bias // Variance Decomposition: SupportVectorClf\n", " Average bias: 0.196\n", " Average variance: 0.007\n", " Average expected loss: 0.197 \"Goodness\": 0.803\n", "\n", "Bias // Variance Decomposition: GaussianNB\n", " Average bias: 0.591\n", " Average variance: 0.081\n", " Average expected loss: 0.575 \"Goodness\": 0.425\n", "\n", "Bias // Variance Decomposition: K-NNeighbors\n", " Average bias: 0.173\n", " Average variance: 0.017\n", " Average expected loss: 0.173 \"Goodness\": 0.827\n", "\n" ] } ], "source": [ "# from our local library\n", "# cross validation 10 folds for faster results\n", "folds = 10\n", "for name, clf in models:\n", " print('Bias // Variance Decomposition:', name)\n", " bias_var_metrics(X_train,X_test,y_train,y_test,clf,folds)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "***" ] }, { "cell_type": "markdown", "metadata": { "tags": [] }, "source": [ "***" ] }, { "cell_type": "markdown", "metadata": { "colab_type": "text", "id": "Cpzuyj7gxwCg", "tags": [] }, "source": [ "**
Imports** for model comparison - CV" ] }, { "cell_type": "code", "execution_count": 27, "metadata": { "tags": [] }, "outputs": [], "source": [ "\n", "# For imbalanced classification, the accuracy score is often uninformative\n", "\n", "# average for each label weighted by support \n", "# balanced accuracy is suitable for model comparison and for use in imbalanced dataset \n", "from sklearn.metrics import balanced_accuracy_score\n", "from sklearn.metrics import f1_score\n", "\n", "eval_metric = ['wtd.avg.accuracy', \n", " 'balanced_accuracy', \n", " 'wtd.avg.f1_score', \n", " 'f1_weighted']\n", "\n", "# for graphs\n", "import matplotlib.pyplot as plt" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "***" ] }, { "cell_type": "code", "execution_count": 28, "metadata": { "tags": [] }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# display confusion matrix of random forest\n", "import matplotlib.pyplot as plt\n", "from scikitplot.metrics import plot_confusion_matrix as splot_cm\n", "splot_cm(y_test, results[0][1], normalize=False)\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": 29, "metadata": { "tags": [] }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# display confusion matrix of LogReg\n", "import matplotlib.pyplot as plt\n", "from scikitplot.metrics import plot_confusion_matrix as splot_cm\n", "splot_cm(y_test, results[1][1], normalize=False)\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": 30, "metadata": { "tags": [] }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# display confusion matrix of support vector\n", "import matplotlib.pyplot as plt\n", "from scikitplot.metrics import plot_confusion_matrix as splot_cm\n", "splot_cm(y_test, results[2][1], normalize=False)\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": 31, "metadata": { "tags": [] }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# display confusion matrix of Gaussian NB\n", "import matplotlib.pyplot as plt\n", "from scikitplot.metrics import plot_confusion_matrix as splot_cm\n", "splot_cm(y_test, results[3][1], normalize=False)\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": 32, "metadata": { "tags": [] }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# display confusion matrix of K-NNeigbor\n", "import matplotlib.pyplot as plt\n", "from scikitplot.metrics import plot_confusion_matrix as splot_cm\n", "splot_cm(y_test, results[4][1], normalize=False)\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": 33, "metadata": { "tags": [] }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# plot ROC curve to display the performance across each attack category \n", "# ROC curve for random forest\n", "from scikitplot.metrics import plot_precision_recall, plot_roc\n", "\n", "clf = models[0][1]\n", "clf.fit(X_train, y_train)\n", "probas = clf.predict_proba(X_test)\n", "\n", "plot_roc(y_test, probas)\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": 34, "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# visualize class balance\n", "from yellowbrick.target import ClassBalance\n", "\n", "# Instantiate the visualizer\n", "visualizer = ClassBalance()\n", "visualizer.fit(y_train, y_test) # Fit the data to the visualizer\n", "_ = visualizer.show() # Finalize and render the figure\n", "# assign visualizer.show() to a null variable to avoid printing some trash" ] }, { "cell_type": "markdown", "metadata": { "colab_type": "text", "id": "Cpzuyj7gxwCg", "tags": [] }, "source": [ " ***" ] }, { "cell_type": "markdown", "metadata": { "colab_type": "text", "id": "Cpzuyj7gxwCg" }, "source": [ "***\n", "**Model Comparison: Cross Validation**" ] }, { "cell_type": "code", "execution_count": 35, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "KFold CV: 5 folds with scoring = balanced_accuracy \n", "\t timer started\n", "RandomForest:\tbalanced_accuracy = 0.852 +/- (0.022)\n", "LogReg" ] }, { "name": "stderr", "output_type": "stream", "text": [ "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ ":\tbalanced_accuracy = 0.756 +/- (0.010)\n", "SupportVectorClf:\tbalanced_accuracy = 0.795 +/- (0.012)\n", "GaussianNB:\tbalanced_accuracy = 0.566 +/- (0.026)\n", "K-NNeighbors:\tbalanced_accuracy = 0.875 +/- (0.026)\n", "\tRun Time 387.22 seconds\n", "\n" ] } ], "source": [ "from sklearn.model_selection import StratifiedKFold, cross_val_score\n", "# a variation of KFold that returns stratified folds\n", "# preserving the percentage of samples for each class\n", "\n", "# ---- Cross Validation ---- #\n", "# number of rounds\n", "folds = 5\n", "\n", "# choose a metric\n", "scorer = eval_metric[1]\n", "# ---- ---- #\n", "\n", "# define the strategy\n", "# Randomly shuffles the data before splitting\n", "# Sets a random seed for reproducibility\n", "skf = StratifiedKFold(shuffle=True, random_state = 11, n_splits=folds)\n", "\n", "from time import time\n", "trs = time()\n", "print('KFold CV: %i folds with scoring = %s \\n\\t timer started' % (folds, scorer))\n", "\n", "results = []\n", "names = []\n", "# Performs cross-validation on the current model using cross_val_score\n", "# Print the average performance with standard deviation.\n", "for name, model in models:\n", " print(name, end='') # no newline at the end\n", " cv_results = cross_val_score(model, X_train, y_train, cv=skf, scoring=scorer)\n", " results.append(cv_results)\n", " names.append(name)\n", " msg = \":\\t%s = %0.3f +/- (%0.3f)\" % (scorer, \n", " cv_results.mean(), \n", " cv_results.std())\n", " print(msg)\n", "\n", "tre = time() - trs\n", "print (\"\\tRun Time {} seconds\".format(round(tre,2)) + '\\n')" ] }, { "cell_type": "code", "execution_count": 36, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "The box shows the middle 50 percent of the data,\n", "the orange line in the middle of each box shows the median of the sample,\n", "and the green triangle in each box shows the mean of the sample.\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ "['RandomForest', 'LogReg', 'SupportVectorClf', 'GaussianNB', 'K-NNeighbors']\n" ] } ], "source": [ "# boxplot model comparison\n", "# One box and whisker plot for each algorithm’s sample of results.\n", "\n", "ttl = scorer + ' Comparison'\n", "print(\"The box shows the middle 50 percent of the data,\\n\"\n", " \"the orange line in the middle of each box shows the median of the sample,\\n\"\n", " \"and the green triangle in each box shows the mean of the sample.\")\n", "\n", "# customizing boxplot appearance\n", "fig = plt.figure()\n", "fig.suptitle(ttl)\n", "ax = fig.add_subplot(111)\n", "plt.boxplot(results, showmeans=True)\n", "#ax.set_xticklabels(names)\n", "plt.show()\n", "print(names)" ] }, { "cell_type": "code", "execution_count": 37, "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stderr", "output_type": "stream", "text": [ "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# individual model per-fold results\n", "# Yellowbrick is a visualization library that provides tools for machine learning models.\n", "# cv_scores function creates a visualization of the cross-validation scores for the models.\n", "from yellowbrick.model_selection import cv_scores\n", "for name, model in models:\n", " viz = cv_scores(model, X_train, y_train, cv=skf, scoring=scorer)" ] }, { "cell_type": "markdown", "metadata": { "colab_type": "text", "id": "Cpzuyj7gxwCg" }, "source": [ "***\n", "**Model Comparison: Bias - Variance Decomposition**" ] }, { "cell_type": "code", "execution_count": 38, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Bias-Variance: Model Comparison \n", "\t timer started\n", "RandomForest: Bias: 0.129 Variance: 0.016 E.loss: 0.134\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n", "D:\\miniconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:460: ConvergenceWarning: lbfgs failed to converge (status=1):\n", "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n", "\n", "Increase the number of iterations (max_iter) or scale the data as shown in:\n", " https://scikit-learn.org/stable/modules/preprocessing.html\n", "Please also refer to the documentation for alternative solver options:\n", " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n", " n_iter_i = _check_optimize_result(\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "LogReg: Bias: 0.227 Variance: 0.008 E.loss: 0.227\n", "SupportVectorClf: Bias: 0.198 Variance: 0.008 E.loss: 0.198\n", "GaussianNB: Bias: 0.569 Variance: 0.084 E.loss: 0.542\n", "K-NNeighbors: Bias: 0.174 Variance: 0.020 E.loss: 0.172\n", "\tRun Time 1365.46 seconds\n", "\n" ] } ], "source": [ "from mlxtend.evaluate import bias_variance_decomp\n", "## bias_variance_decomp() requires \n", "## 1. numpy ndarrays\n", "## 2. numeric targets\n", "from sklearn.preprocessing import LabelEncoder\n", "\n", "\n", "folds = 12\n", "\n", "from time import time\n", "trs = time()\n", "print('Bias-Variance: Model Comparison \\n\\t timer started')\n", "\n", "#Encode target labels to numerical values\n", "#because bias_variance_decomp function requires numerical target values\n", "ytrain = LabelEncoder().fit_transform(y_train)\n", "ytest = LabelEncoder().fit_transform(y_test)\n", "\n", "# for graphs\n", "rx = np.arange(len(models)) # the x locations for the groups\n", "\n", "#Calculate the average expected loss, bias, and variance for the current model.\n", "# Sets a random seed for reproducibility\n", "cn, bias, var, err = [], [], [], []\n", "for name, clf in models:\n", " avg_expected_loss, avg_bias, avg_var = bias_variance_decomp(\n", " clf, X_train.values, ytrain, X_test.values, ytest, \n", " loss='0-1_loss', num_rounds=folds, random_seed=150)\n", " err.append(avg_expected_loss)\n", " bias.append(avg_bias)\n", " var.append(avg_var)\n", " cn.append(name)\n", " \n", " print(name,end='') \n", " msg=\": Bias: %0.3f Variance: %0.3f E.loss: %0.3f\" % (avg_bias, avg_var, avg_expected_loss)\n", " print(msg)\n", " \n", "tre = time() - trs\n", "print (\"\\tRun Time {} seconds\".format(round(tre,2)) + '\\n')" ] }, { "cell_type": "code", "execution_count": 39, "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ "['RandomForest', 'LogReg', 'SupportVectorClf', 'GaussianNB', 'K-NNeighbors']\n" ] } ], "source": [ "#visualisations\n", "# line plot\n", "fig, ax = plt.subplots()\n", "ax.plot(rx, err, 'b', label = 'total_error')\n", "ax.plot(rx, bias, 'k', label = 'bias')\n", "ax.plot(rx, var, 'y', label = 'variance')\n", "ax.legend()\n", "ax.set_xticks(rx)\n", "ax.set_title('Bias-Variance Decomposition')\n", "plt.show()\n", "print(cn)" ] }, { "cell_type": "code", "execution_count": 40, "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ "['RandomForest', 'LogReg', 'SupportVectorClf', 'GaussianNB', 'K-NNeighbors']\n" ] } ], "source": [ "# stacked bar plot which reflects better on expected loss, bias and variance\n", "\n", "fig = plt.figure()\n", "ax = fig.add_axes([0,0,1,1])\n", "ax.bar(rx, err, color = 'b', width= 0.35)\n", "ax.bar(rx, bias, color = 'g', width= 0.35)\n", "ax.bar(rx, var, color = 'r', width= 0.35)\n", "ax.legend(labels=['E.loss', 'Bias', 'Var'])\n", "ax.set_title('Bias-Variance Decomposition')\n", "ax.set_xticks(rx)\n", "plt.show()\n", "print(cn)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "tags": [] }, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.8.19" } }, "nbformat": 4, "nbformat_minor": 4 }