Modified genetic algorithm based feature subset selection in intrusion detection system

Feature subset selection is important not only for decreasing computing spending but also for the improved understandability and accuracy of the classification results. In this paper, we proposed a combined feature subset selection method, called RICGA (ReliefF immune clonal genetic algorithm), based on the ReliefF algorithm, immune clonal selection algorithm and GA. In the RICGA method, we first use ReliefF to get rid of irrelevant features, then apply a modified genetic algorithm to acquire the finally feature subset. We analyze roughly the Markov chain model of RICGA algorithm and its convergence. The experimental results on real KDD CUP´99 dataset show that the RICGA method is superior to the GA and ReliefF-GA on classification accuracy and accepted feature subset size.