An Effective Metaheuristic Algorithm for Intrusion Detection System

Abstract

Intrusion detection system (IDS) is typically used to detect and prevent abnormal behaviors in a network management system. The basic idea of IDS is to use feature values from network packets capture mechanism to classify whether a behavior is abnormal. However, most traditional classification algorithms are incapable of recognizing unknown behaviors. To develop a high performance classification algorithm to improve the accuracy of IDS, the algorithm proposed in this paper will integrate clustering, classification, and metaheuristic algorithms as a classification algorithm for IDS, called search economics with k-means and support vector machine (SEKS). Moreover, this hybrid strategy for the proposed algorithm is aimed at improving the accuracy of abnormal behavior detection of such a system, reducing the computation time of a classification algorithm, and making it possible for the IDS to recognize the unknown and new variant attacks in a network environment. The experimental results show that the proposed algorithm outperforms all the other classification algorithms compared in this paper in terms of the accuracy.