Intelligent intrusion detection system using fuzzy rough set based C4.5 algorithm

Abstract

In recent years, as the usage of internet increases, new type of attacks on network information is also increasing continuously. Intrusion Detection System (IDS) is an important component that provides security to the network information by identifying various kinds of attacks occurring in the networks. Currently, there are many researches who are working in this area and they focus on developing effective IDS using machine learning techniques. However, there is a need for better systems with improved detection accuracy and reduced false alarm rate. In this paper, we propose an Intelligent IDS using fuzzy rough set based C4.5 classification algorithm to improve the detection accuracy. This system has been compared with Support Vector Machines for illustrating the improvement with respect to the detection accuracy. The inputs to these classifiers were preprocessed using a fuzzy rough set based outlier detection algorithm. In this work, we used the KDD'99 Cup dataset for carrying out the simulation of the experiments. The experimental results obtained in this work show that the proposed model reduces the false alarm rate and improves overall detection accuracy.