Network Intrusion Detection Model Based on Least Squares Support Vector Machine Optimized by Modify Cuckoo Search Algorithm

In order to improve detection rate of the network intrusion, this paper proposes a intrusion detection model (MCS-LSSVM)which the least square support vector machine (LSSVM) is optimized by improved cuckoo search(MCS) algorithm.Firstly,the parameters of LSSVM are taken as a nest location of cuckoo,and then the cuckoo species parasitic generation mechanism is used to find the optimal nest and transformed into LSSVM optimal parameters,finally,the optimal parameters is used to establish intrusion detection models.The simulation results show that the proposed model not only improves the detection rate of network intrusion,and shorten the training time to improve the network intrusion detection effect.