JAYA based Classifier System for Cybersecurity

Cybersecurity incidents, whether due to human actions or natural disasters, are becoming a significant threat to businesses, organizations, and governments. The frequency of cyber-attacks is increasing, and these attacks can spread quickly through platforms such as flash devices and hard disks. Protecting a network from such threats is the foremost priority, and an intrusion detection system (IDS) plays a key role. An IDS monitors networks and analyzes data to identify abnormal or malicious activity. This study proposes a new method for strengthening our defense against cybersecurity threats by integrating the Jaya algorithms with classification systems. We used optimization techniques to increase the efficiency of these models in countering cyberattacks by applying JAYA algorithms instead of known and regular genetic units, that is, genetic algorithms. We introduce a Jaya-based framework that optimizes the breeding classifier, namely, the feed classification system, to improve detection accuracy. The primary experimental results indicate that the proposed work's detection rate (DR) exceeds 90% with a false alarm rate of less than 5%. All experiments were conducted on the UNSW-NB15 dataset to demonstrate the detection efficiency.