Application of Improved Butterfly Optimization Algorithm Combined with Black Widow Optimization in Feature Selection of Network Intrusion Detection

Feature selection is a very important direction for network intrusion detection. However, current feature technology of detection has the problems low rate and accuracy due to redundancy. An improved Butterfly Optimization Algorithm combined with Black Widow (BWO-BOA) proposed in this paper, which introduces dynamic adaptive search strategy global phase (BOA), uses movement process (BWO) algorithm as local search, at same time, order overcome butterfly optimization easily falling into optimum phase, takes advantage small probability mutation filter out redundant features. This paper then tries apply BWO-BOA In verify performance algorithm, UNSW-NB15 dataset selected binary classification multi-classification simulation experiments, models BOA BWO Particle Swarm Optimization, Salp Algorithm, Whale are compared validation. The experimental results show that can enhance model significantly boost reduction dimensions.