Document Type: Original Manuscript
Department of Computer Engineering, Kerman Branch, Islamic Azad University, Kerman, Iran
Intrusion detection systems are designed to provide security in computer networks, so that if the attacker crosses other security devices, they can detect and prevent the attack process. One of the most essential challenges in designing these systems is the so called curse of dimensionality. Therefore, in order to obtain satisfactory performance in these systems we have to take advantage of appropriate pre-processing steps specially the feature subset selection methods. Since the problem of searching for the optimal feature subset has an intolerable complexity, in this paper we propose a genetic-algorithm-based search method for finding the most relevant subset of features. In order to find the most relevant features, the parallel structure of the genetic algorithm along with the distribution factor of the features is used. The fitness value of each feature subset is computed according to performance of the classifier trained with respect to that subset. In order to evaluate the performance of the proposed method, we use the NSL-KDD dataset which benefits from more real-world intriguing records than other intrusion detection data. The results of our evaluation experiments shows that the proposed method outperforms the prior methods.