Improving the Particle Swarm Algorithm and Optimizing the Network Intrusion Detection of Neural Network

摘要

According to the connections between the feature subset and RBF neural network parameter, in order to improve the accuracy rate of intrusion detection, a network intrusion detection model ( IPSO-BPNN ) improving the article swarm and optimizing the neural network is put forward. Take the network feature subset and RBF neural network parameter as a particle, and discover the optimum network feature subset and RBF neural network parameter through the coordination and information exchange between particles to establish the optimum network intrusion detection model, and adopt KDD Cup99 data set to perform the simulation experiment. The results of simulation experiment show that, IPSO-RBF neural network reduces the feature dimensions, and obtains better RBF neural network parameter, which is a network intrusion detection model of high detection accuracy rate and speed.