Abstract:With the continuous occurrence of cybersecurity incidents, network intrusion detection has become one of the most critical issues in cyber ecosystems. Although previous machine learning-based approaches have made significant progress, their generalization ability is limited due to the following critical challenges. First, intrusion detection is severely affected by the class imbalance problem in many network scenarios, with some attack types representing only a very small subset of the entire training set. Second, cyberattacks are becoming increasingly sophisticated, and hence, it is becoming more challenging for existing methods to extract robust representations. Third, most existing methods generally leverage only a particular aspect of the network traffic features and treat model training as a single-task learning problem, thus ignoring the discriminative ability of different feature types and the performance enhancement of integrating multiple machine learning tasks. In this paper, we propose a Multi-task lEarning Model with hyBrid dEep featuRes (MEMBER) to address the aforementioned challenges. Based on a Convolutional Neural Network (CNN) with embedded spatial and channel attention mechanisms, MEMBER innovatively introduces two auxiliary tasks (i.e., an auto-encoder (AE) enhanced with a memory module and a distance-based prototype network) to boost the model generalization ability and alleviate the performance degradation suffered in imbalanced network environments. Extensive experiments on several benchmark datasets demonstrate the superiority and robustness of our proposed MEMBER in terms of both F1 score and stability.

Intrusion Detection Based on Multi-Layer Minimax Probability Machine Classifier

A Mixed Intrusion Detection System Utilizing K-means and Extreme Gradient Boosting

A Novel Intrusion Detection System Based on an Optimal Hybrid Kernel Extreme Learning Machine

Multi-agent based hybrid Intrusion detection system

Transductive Minimax Probability Machine.

Hybrid Intrusion Detection Method Based On K-Means And Cnn For Smart Home

A BOOSTING APPROACH FOR INTRUSION DETECTION

A Comparative Study of Minimax Probability Machine-based Approaches for Face Recognition.

Network intrusion detection by using multiple-criteria linear programming

Network Intrusion Detection by Multi-group Mathematical Programming based Classifier

Open-Set Intrusion Detection with MinMax Autoencoder and Pseudo Extreme Value Machine

MSML: A Novel Multilevel Semi-Supervised Machine Learning Framework for Intrusion Detection System.

Error-correcting Ability Based Collaborative Multi-Layer Selective Classifier Ensemble Model for Intrusion Detection

Building an Effective Intrusion Detection System Using the Modified Density Peak Clustering Algorithm and Deep Belief Networks

Hybrid Intrusion Detection Method Based on CM-K-Means

A Network Intrusion Detection Model Based on BiLSTM with Multi-Head Attention Mechanism

A supervised machine learning-based solution for efficient network intrusion detection using ensemble learning based on hyperparameter optimization

A Multi-Channel Contrastive Learning Network Based Intrusion Detection Method

An Incremental Majority Voting Approach for Intrusion Detection System Based on Machine Learning

MEMBER: A multi-task learning model with hybrid deep features for network intrusion detection

The Use of Ensemble Models for Multiple Class and Binary Class Classification for Improving Intrusion Detection Systems