Intrusion Detection Model Based On Chicken Swarm Optimization Algorithm And Kernel Limit Learning Machine

With the rapid development of network technology and the increasing popularity of network applications,network security problem has also started to receive much attention.The "Network Security Trends Report for the First Half of 2019" shows that the number of network attacks is increasing year by year,and the main target of network attacks is in the corporate and education fields.Network attack not only may lead to the theft of user's information,but also cause damage to computer and industrial system in severe cases.Research on network attack detection methods and technologies is urgent.Intrusion detection model detects attack events by analyzing security log,network traffic and other information.However,due to the massiveness and high dimensionality of the data,Intrusion detection model cannot make timely and accurate judgments on the attack behavior.In view of the above problems,this paper proposes an intrusion detection model based on the improved chicken swarm optimization algorithm and the kernel extreme learning machine.The ICSO algorithm is used to simultaneously handle the feature selection of intrusion detection dataset and parameter optimization of the KELM classifier,try to reduce the computing resources occupied by the model on the basis of ensuring the detection performance of the model.The main work of this paper is as follows:(1)An improved swarm optimization algorithm(ICSO)is proposed.Considering that the chicken swarm optimization algorithm is easy to fall into the local optimal solution when processing the data with high dimension,in order to speed up the convergence rate of the algorithm and enhance the ability of the algorithm to jump out of the local optimal solution,the following improvements are made to the chicken swarm optimization algorithm.Firstly,the chaos sequence and reverse learning are used to initialize the population,and the initial chicken population with uniform distribution and diversity is obtained.Then introduce a non-linear inertia weight factor into the hen's position update formula,and make the hen learn from the rooster in the group while learning from the chickens with the best fitness value in the swarm.Finally,when the algorithm converges prematurely,Cauchy mutation is introduced to give the algorithm the ability to jump out of the local optimal solution.(2)An intrusion detection model(ICSO-KELM)based on the synchronous optimization of KELM parameters and data set characteristics of the ICSO algorithm is proposed.Considering the adaptation of feature selection and classifier parameter optimization,this paper aims to reduce the feature dimension of the dataset and improve the accuracy of the classifier.The ICSO algorithm is used to synchronize the feature selection of the dataset and the optimization of the KELM kernel parameters.Find the optimal kernel parameters of KELM while optimizing the subset of features.(3)The ICSO-KELM model is simulated in KDD-CUP99 data set.The data after preprocessing is segmented by using 10-cv method,and the data samples after feature selection are trained and detected by optimized KELM.The experimental results show that compared with the PSO-KELM model,GA-KELM model,and CSO-KELM model,the ICSO-KELM model obtains fewer feature numbers and better detection accuracy;compared withclassic machine learning such as SVM and BP Algorithm,ICSO-KELM model improves the accuracy of intrusion detection and reduces the false alarm rate of intrusion detection to a certain extent.