Enhancing Feature Selection in Network Intrusion Detection Systems Using a Novel Hybrid Binary Swarm Algorithms
Maher Khalaf Hussein, Asmaa Alqassab, Lubna Thanoon Alkahla · Baghdad Science Journal · 2025
An intrusion detection system (IDS) is a system that monitors network traffic for any mistrustful activity and issues alerts when it is detected. The selection of feature stage is important in effective intrusion detection systems, as it determines the features that are most influential in detecting normal traffic and malicious traffic. In this study, new proposed hybrid method for selecting features in intrusion detection systems based on the hybrid binary Salp Swarm algorithm (SSA) and the binary Pigeon Inspired Optimizer (PIO), inspired by the collective behavior of animals. The proposed method removes unnecessary features and increases the number of features necessary, as the classification accuracy reached 99 percent. This feature selection approach increases the overall performance of intrusion detection systems by combining the strengths of binary SSA and PIO.