An effective intrusion detection method using optimal hybrid model of classifiers

Abstract

With increasing connectivity between computers, the security of computer networks plays a strategic role in modern computer systems. In order to enforce high protection levels against threats, a number of software systems have been currently developed. Intrusion detection systems (IDS) have become an essential component at detecting intruders. In this paper, an ensemble approach to network intrusion detection based on the fusion of multiple classifiers is proposed. A computational machine is built to derive optimal parsimonious hybrid model of classifiers in intrusion detection based on the following classification methods, Naive Bayes, Support Vector Machine, K-nearest neighbor, and Neural networks. The weighted voting fusion strategy for intrusion detection is assessed by experiments and its performances compared. The potentialities of classifiers fusion for the development of effective intrusion detection systems are evaluated and discussed. The experimental results indicate that hybrid approach effectively generates a more accurate intrusion detection model on detecting both normal usages and malicious activities. In this paper, we aim to build a robust classifier combination system given a classifier set.