Multi-population Diversity-guided Genetic Algorithm for Feature Selection in Network Intrusion Detection

2026-05-20 · Source: cs.NE updates on arXiv.org · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cybersecurity & Data Privacy · Depth: Expert, extended

Summary

A new Multi-Population Diversity-Guided Genetic Algorithm (MPDGGA) has been developed to enhance feature selection in Network Intrusion Detection Systems (NIDS). This algorithm addresses common limitations in existing Genetic Algorithm-based methods, such as difficulty maintaining population diversity and a lack of guidance for evolutionary operators in high-dimensional, redundant traffic features. MPDGGA introduces a chained multi-population evolutionary structure and a diversity-guided operator based on the information gain ratio. Experimental results across 11 datasets, including NSL-KDD, UNSW-NB15, and 9 UCI datasets, demonstrate that MPDGGA significantly outperforms four other advanced multi-population feature selection models. It achieved the highest accuracy on 10 of these datasets, selecting at least 2.26% of features, and notably reduced the feature selection ratio by 9.76% to 19.51% on NSL-KDD compared to the second-best model.

Key takeaway

For research scientists developing NIDS, MPDGGA offers a robust approach to feature selection that mitigates premature convergence and diversity loss. You should consider implementing its chained multi-population structure and diversity-guided operators to achieve higher classification accuracy with more compact feature subsets, especially in high-dimensional traffic data. This can lead to more efficient and effective intrusion detection systems.

Key insights

MPDGGA improves NIDS feature selection by combining multi-population evolution with diversity-guided operators for better accuracy and feature compactness.

Principles

• Chained multi-population structures enhance global and local exploration.
• Information gain ratio can pre-evaluate feature subsets efficiently.
• Diversity-guided operators improve convergence and maintain population diversity.

Method

MPDGGA uses binary encoding, partitions populations into a chained structure, and applies diversity-guided crossover and mutation operators based on an information gain ratio subset evaluation criterion to optimize feature selection.

In practice

• Apply MPDGGA for NIDS to reduce feature redundancy.
• Use information gain ratio for quick feature subset pre-evaluation.
• Consider multi-population structures for high-dimensional feature spaces.

Topics

• Network Intrusion Detection Systems
• Feature Selection
• Multi-Population Genetic Algorithm
• Diversity-guided Operators
• Information Gain Ratio

Best for: Research Scientist, AI Scientist, Machine Learning Engineer, AI Security Engineer

Related on AIssential

• MOELIGA: a multi-objective evolutionary approach for feature selection with local improvement — MOELIGA refines feature selection using multi-objective genetic algorithms with local improvement and diverse fitness sharing.
• An adaptive subsampling method for large-sample feature screening — BanditCR-SIS accelerates model-free feature screening using multi-armed bandits and Chatterjee's rank correlation, drastically cutting computational costs.
• Mathematical perspective on genetic algorithms with optimization guided operators — ML-guided genetic algorithm operators, though costly, improve optimization by leveraging diversity, modeled as a query-complexity problem.
• Cyber-Physical Anomaly Detection in IoT-Enabled Smart Grids Using Machine Learning and Metaheuristic Feature Optimization — Optimized feature selection significantly enhances cyber-physical anomaly detection in smart grids, reducing data redundancy while improving accuracy.
• Polynomial-Time Optimal Group Selection via the Double-Commutator Eigenvalue Problem — This paper introduces a polynomial-time algorithm for the "group selection" problem within the algebraic diversity (AD) framework, which replaces temporal averaging with algebraic group action for sec…
• A Hybrid Approach For Malware Classification Using Secondary Features Fusion — A hybrid method combines feature and algorithm fusion for automated, accurate malware family classification.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by cs.NE updates on arXiv.org.