Anomaly Detection Algorithm Based on FCM with improved Krill Herd

Abstract

The anomaly detection algorithm plays an important role in the field of network security. Among them, the most representative method is anomaly detection based on fuzzy C-means (FCM). FCM relies heavily on the initial clustering center and is prone to local extremum. Therefore, the detection effect of the FCM-based anomaly detection algorithm is not ideal in some cases. The herd intelligent optimization technology has a strong global search capability and is widely used in various fields. As a herd intelligence technology, the krill herd algorithm has a relatively simple optimization function structure, which has strong global search ability and is easy to integrate with other optimization strategies. Therefore, a KH algorithm with strong global search capability is introduced, and a hybrid KH-FCM algorithm is proposed. In the hybrid KH-FCM algorithm, the randomly generated initial population will be divided into two subpopulations containing the same number of individuals.