Abstract:The contributions of the paper are as follows: The introduction of an ensemble learning‐based stacking classifier for anomaly detection in communication network traffic. The utilization of base classifiers such as K‐nearest neighbour (KNN), logistic regression (LR), Naive Bayes (NB), and decision tree (DT), with Support Vector Machines (SVM) as a meta‐classifier for anomaly detection. A comprehensive comparison of accuracy, precision, recall, and F‐measure between the proposed model and individual models, including KNN, NB, DT, LDA, LR, and SVM. In recent years, the internet has not only enhanced the quality of our lives but also made us susceptible to high‐frequency cyber‐attacks on communication networks. Detecting such attacks on network traffic is made possible by intrusion detection systems (IDS). IDSs can be broadly divided into two groups based on the type of detection they provide. According to the established rules, the first signature‐based IDS detects threats. Secondly, anomaly‐based IDS detects abnormal conditions in the network. Various machine and deep learning approaches have been used to detect anomalies in network traffic in the past. To improve the detection of anomalies in network traffic, researchers have compared several machine learning models, such as support vector machines (SVM), logistic regressions (LRs), K‐Nearest Neighbour (KNN), Nave Bayes (NBs), and boosting algorithms. The accuracy, precision, and recall of many studies have been satisfactory to an extent. Therefore, this paper proposes an ensemble learning‐based stacking classifier (ELSC) to achieve a better accuracy rate. In the proposed ELSC algorithm, KNN, NB, LR, and Decision Trees (DT) served as the base classifiers, while SVM served as the meta classifier. Based on a Network Intrusion detection dataset provided by Kaggle.com, ELSC is compared to base classifiers such as KNN, NB, LR, DT, SVM, and Linear Discriminate Analysis. As a result of the simulations, the proposed ELBS stacking classifier was found to outperform the other comparative models and converge with an accuracy of 99.4%.

SALAD: A split active learning based unsupervised network data stream anomaly detection method using autoencoders

An Unsupervised Framework for Anomaly Detection in a Water Treatment System

SALAD: Self-Adaptive Lightweight Anomaly Detection for Real-time Recurrent Time Series

A network anomaly detection algorithm based on semi-supervised learning and adaptive multiclass balancing

SLADE: Detecting Dynamic Anomalies in Edge Streams without Labels via Self-Supervised Learning

Stream-based Active Learning for Anomalous Sound Detection in Machine Condition Monitoring

Boundary-Driven Active Learning for Anomaly Detection in Time Series Data Streams.

Anomaly Detection of Network Streams Via Dense Subgraph Discovery

Big data analysis and distributed deep learning for next-generation intrusion detection system optimization

HELAD: A novel network anomaly detection model based on heterogeneous ensemble learning

Learning to Detect: A Data-driven Approach for Network Intrusion Detection

A hybrid IndRNNLSTM approach for real-time anomaly detection in software-defined networks

Autoencoder-based Unsupervised Intrusion Detection using Multi-Scale Convolutional Recurrent Networks

Unsupervised Anomaly Detection in Spatio‐Temporal Stream Network Sensor Data

Unsupervised real-time anomaly detection for streaming data

Unsupervised Anomaly Detection in Stream Data with Online Evolving Spiking Neural Networks

Self-Supervised Learning for Online Anomaly Detection in High-Dimensional Data Streams

Deep Learning-Based Anomaly Detection in Network Traffic for Cyber Threat Identification

Anomaly detection in network traffic with ELSC learning algorithm

A Novel Two-Stage Deep Learning Structure for Network Flow Anomaly Detection

Anomaly Intrusion Detection Based on Data Stream