Abstract:

Because fuzzy Cmeans clustering method is sensitive to initial cluster centers and easily trapped into local minima, we cant get precise classification result in network intrusion detection. To solve the problem, a network intrusion detection method based on GSO(Glowworm Swarm Optimization) algorithm is proposed. First, samples with label is used to get initial cluster center. Then, GSO is employed to optimize cluster center. Simulation result shows that the method is effective.

Key words: glowworm swarm optimization (GSO) algorithm, network intrusion, fuzzy C-means clustering, semi-supervised