Title: Improving ID performance using GA and NN

Authors: S. Selvakani, R.S. Rajesh

Addresses: Department of Computer Applications, PSN College of Engineering and Technology, Tirunelveli – 627 152, Tamilnadu, India. ' Department of Computer Science and Engineering, Manonmaium Sundaranar University, Tirunelveli – 627 012, Tamilnadu, India

Abstract: The internet has been growing at an amazing rate and concurrent with the growth, the vulnerability is also increasing. How to find and detect novel or unknown attacks is one of the most important objectives in current IDS. Most of the current IDS examine all data features to detect intrusions. However, some of the features may be redundant or contribute little to the detection process. This paper mainly addresses the issue of identifying important input features for intrusion detection. This paper proposes an intrusion detection model that is computationally efficient and effective based on mutual information. Then genetic algorithm is applied to generate optimal rules. Those generated rules are used to detect known attacks. RBF is also used to learn and detect unknown attacks. Experimental results on the well-known KDD 99 data set show the achievement of high true positive rates and acceptable low false positive rates and are effective.

Keywords: anomaly detection; confusion matrix; genetic algorithms; GAs; information gain; knowledge synthesis; radial basis function; RBF neural networks; internet intrusions; internet attacks; intrusion detection.

DOI: 10.1504/IJCAET.2008.021257

International Journal of Computer Aided Engineering and Technology, 2008 Vol.1 No.1, pp.81 - 93

Published online: 14 Nov 2008 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article