A Parallel Clustering Ensemble Algorithm for Intrusion Detection System

Clustering analysis is a common unsupervised anomaly detection method, and often used in Intrusion Detection System (IDS), which is an important component in the network security. The single cluster algorithm is difficult to get the great effective detection, and then a new cluster algorithm based on evidence accumulation is born. The IDS with clustering ensemble has a low false positive rate and high detection rate, however, the IDS is slow to detect the mass data stream, and it can not detect the attacks in time. This paper presents a parallel clustering ensemble algorithm to improve the speed and the effective of the system. Finally, the KDDCUP99 data set is used to test the system show that the IDS have greatly improvement in time and efficiency.