Best hybrid classifiers for intrusion detection

Issue title: Selected papers from the International Conference on Computer Science, Software Engineering, Information Technology, e-Business, and Applications, 2004

Guest editors: Narayan C. Debnath

Article type: Research Article

Authors: Kholfi, Sanaa^{a; *} | Habib, Muhammad^a | Aljahdali, Sultan^b

Affiliations: [a] School of Information Technology, George Mason University, Fairfax, VA 22033, USA | [b] College of Business Administration (CBA), P.O. Box 110200, Jeddah 21361, Saudi Arabia

Correspondence: [*] Corresponding author. E-mail: skholfi@gmu.edu.

Abstract: We present in this paper an intrusion detection software-system that we have built based on combined statistical and computational models to detect intrusions and classify them as attack or non-attack. More specifically, we build a computational machine to derive optimal parsimonious hybrid model of classifiers in intrusion detection. The classifiers are based on the following classification methods, Naïve Bayes-NB, K-nearest neighbor-K-nn, and Neural networks-NN.

Keywords: Intrusion detection, optimal hybrid model, network security, Naïve Bayes-NB, K-nearest neighbor-K-nn, and Neural networks-NN

DOI: 10.3233/JCM-2006-6S208

Journal: Journal of Computational Methods in Sciences and Engineering, vol. 6, no. s2, pp. S299-S307, 2006