Abstract—The escalation of cyber threats in large-scale local area networks necessitate advanced strategies for efficient anomaly detection and intrusion prevention. This paper explores the integration of sophisticated machine learning techniques and feature selection methods to enhance the performance of Network Intrusion Detection Systems. Focusing on the complex landscape of cyber threats, accentuated by the proliferation of technologies such as Internet of Things, 5G, and cloud computing, the proposed study evaluates the application of three advanced feature selection algorithms—Grey Wolf Optimizer, Bat Algorithm, and Pigeon-inspired Optimization—to identify an optimal subset of features that accurately differentiate between diverse cyberattacks and normal network traffic. Employing the comprehensive CSE-CIC-IDS2018 dataset, the experimental results demonstrate that the feature set was successfully reduced from 80 to subsets of 10, 6, and 7 features while maintaining a high detection accuracy close to 99%. This reduction in feature space significantly decreases computational overhead without compromising detection capability. This research contributes to the cybersecurity domain by presenting a scalable, efficient, and highly accurate model for intrusion detection, setting a foundation for future advancements in Network Intrusion Detection Systems optimization and the broader field of cyber defense mechanisms.
 
Keywords—feature selection algorithm, network intrusion detection systems, machine learning, cybersecurity, anomaly detection, cyber threats


Copyright © 2025 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).