Application research of improved K-means algorithm in network intrusion detection

The original K-means algorithm is sensitive to the selection of the initial clustering center and unstable in the network intrusion detection application based on data mining. In this paper, the optimization ability of particle swarm optimization (PSO) was used to solve the problem that K-means algorithm is sensitive to the selection of the initial clustering center. Different global optimal solutions were obtained by PSO algorithm and used as the basis of choosing the initial clustering center of K-means clustering algorithm. Based on this idea, the K-means algorithm was improved and a network intrusion detection model was established. Experimental results show that the improved K-means clustering algorithm based on PSO has better clustering effect than original K-means clustering algorithm, and can detect more intrusion behaviors in network intrusion detection.