Research of K-PSO Cluster Algorithm in Intrusion Detection
Ming Huang · Science Technology and Engineering · 2009
Based on the improved particle swarm optimization algorithm and the k-mean clustering,a new K-PSO cluster algorithm is proposed.First the improved particle swarm optimization algorithm is used to find the optimal k initial cluster center,then find the cluster result through K-means algorithm,finally output the result.In the algorithm,every vector in the vector space be described as a point;every program in the dataset a dimension in solution space;the whole dataset a multi-dimensional space with many points.Every point will be mapped to a particle.The whole dataset is a particle swarm.The experimental results show that the PSO cluster algorithm can improve the DR and reduce the FAR of Abnormal Detection in intrusion detection system.