No abstract

In today's world, almost everybody is affluent with computers and network based technology is growing by leaps and bounds. So, network security has become very important, rather an inevitable part of computer system. An Intrusion Detection System (IDS) is designed to detect system attacks and classify system activities into normal and abnormal form. Machine learning techniques have been applied to intrusion detection systems which have an important role in detecting Intrusions. This paper reviews different machine approaches for Intrusion detection system. This paper also presents the system design of an Intrusion detection system to reduce false alarm rate and improve accuracy to detect intrusion.

No abstract

Network security is a very important aspect for internet enabled systems. As the internet keeps developing the number of security attacks as well as their severity has shown a significant increase. The Intrusion Detection System (IDS) plays a very important role in discovering anomalies and attacks in the network. The aim of an intrusion detection system is to identify those entities that attempt to destabilize security controls that have been put in place. The field of machine learning is rapidly gaining more attention in the development of these intrusion detection systems. Machine learning techniques can be broadly classified into three broad categories: Supervised, Un-supervised and semi-supervised. The supervised learning method displays good classification accuracy for those attacks that are aready known to us. But this method requires a large amount of training data.The availability of labelled data is not only time consuming but also very expensive. The evolving field of semi-supervised learning offers a promising direction for supplementary research. Hence, in this paper we propose a semi-supervised approach for a pattern based IDS to improve performance and to reduce the false alarm rate. The experimentation is performed on KDD CUP99 dataset and we use the J48 Algorithm in order to implement the semi-supervised learning.

No abstract