Efficient Intrusion Detection using Weighted K-means Clustering and Naïve Bayes Classification

Intrusion detection system (IDS) is becoming a vital component to secure the network. A successful intrusion detection system requires high accuracy and detection rate. In this paper a hybrid approach for intrusion detection system based on data mining techniques is proposed. The principal ingredients of the approach are weighted k-means clustering and naive bayes classification. The C5.0 algorithm is used for ranking attributes, so the attributes receive a weight which is used in K-means clustering therefore accuracy of clustering is increased