Application of Rough Set Theory to Intrusion Detection System

In intrusion detection systems, many intelligent information processing methods, soft computing technology and so on have been applied to generating attack signatures automatically, updating signatures easily and improving detection accuracy with ultra data sets. This paper presents a network intrusion detection system based on rough set theory. The system exploits data reductions, rule selection, feature selection of rough set theory to improve detection accuracy, preprocess data and reduce false alarm and unreal alarm. Empirical results illustrate that the intrusion detection model can detect intrusions accurately.