Intrusion Feature Selection Algorithm Based on Particle Swarm Optimization

High-dimensional intrusion detection data concentration information redundancy results in lower processing velocity of intrusion detection algorithm. Accordingly, the current study proposes an intrusion feature selection algorithm based on particle swarm optimization (PSO). Analyzing the features of the relevance between network intrusion data allows the PSO algorithm to optimally search in a featured space and autonomously select effective feature subset to reduce the data dimensionality. Experimental results show that the algorithm can effectively eliminate redundancy and reduce intrusion feature selection time to effectively increase the detection velocity of the system while ensuring detection accuracy rate.