Abstract:
In recent years, computer systems are facing increased number of security threats because of rapid expansion of computer networks. Different soft computing techniques have been proposed in recent years to develop the Intrusion Detection System. This paper presents an effective genetic algorithm (GA) approach for intrusion detection and the software implementation. The Genetic algorithm is used to derive the set of classification rules from audit data and support confidence framework is utilizes as fitness function to judge the quality of each rule. Then the generated rules are used to detect or classify network intrusions. The proposed method is easy to implement while providing the flexibility to either generally detect network intrusions. Experimental results show the more effective detection rates based on benchmark DARPA data sets on intrusions.

Keywords: Genetic algorithm, Intrusion Detection, support confidence framework