An efficient approach for Intrusion Detection using data mining methods

Intrusion Detection System (IDS) is becoming a vital component of any network in today´s world of Internet. IDS are an effective way to detect different kinds of attacks in an interconnected network thereby securing the network. An effective Intrusion Detection System requires high accuracy and detection rate as well as low false alarm rate. This paper focuses on a hybrid approach for intrusion detection system (IDS) based on data mining techniques. The main research method is clustering analysis with the aim to improve the detection rate and decrease the false alarm rate. Most of the previously proposed methods suffer from the drawback of k-means method with low detection rate and high false alarm rate. This paper presents a hybrid data mining approach encompassing feature selection, filtering, clustering, divide and merge and clustering ensemble. A method for calculating the number of the cluster centroid and choosing the appropriate initial cluster centroid is proposed in this paper. The IDS with clustering ensemble is introduced for the effective identification of attacks to achieve high accuracy and detection rate as well as low false alarm rate.