An Improved Grey Wolf Optimizer Algorithm Integrated With Cuckoo Search And Its Application To Intrusion Detection

With the increasing development of network technologies,the network security situation has become increasingly severe,which have received increasing attention from people.As be an important part of network security,intrusion detection technology has always been a hot issue for its optimization.For the characteristics of intrusion detection data sets that have many redundant features,this thesis conducts in-depth research on feature selection methods and simultaneous selection methods of parameters with features in order to improve intrusion detection efficiency.The main research contents are as follows:(1)Grey Wolf Optimizer algorithm integrated with Cuckoo Search(CS-GWO)is proposed.Grey wolf optimization algorithm is easy to fall into a local optimum while applying to a data set with a high dimensionality.Therefore,in order to enhance the ability of the grey wolf optimization algorithm to jump out of the local optimum,CS-GWO algorithm is proposed,which has a better global search ability combined with cuckoo search.(2)Intrusion detection feature selection based on CS-GWO algorithm.The feature dimension of intrusion detection datasets is high,which is easy to cause dimensional disaster problems and has a low detection accuracy if all the features are utilized to train the classification model.At the same time,it will lead to large overhead and inefficiency.In this thesis,the proposed CS-GWO algorithm is applied to the feature selection process of intrusion detection through selecting the optimal feature subset from the original feature set,for the sake of improving the accuracy of intrusion detection and reducing computational overhead during intrusion detection.(3)Simultaneous selection of parameters and features for SVM based on CS-GWO algorithm for intrusion detection.In essence,intrusion detection is a classification problem.Not only the selection of the feature subset in datasets,but also the penalty parameters,kernel function selection and kernel function parameters of the SVM classifier have a great influence on the classification performance.In order to improve the performance of the SVM classifier and intrusion detection effectiveness,the proposed CS-GWO algorithm is applied to simultaneous optimization process,which combines the feature selection and the parameter optimization process of the classifier at the same time.Finally,compared with the classic particle swarm optimization algorithm,grey wolf optimization algorithm,and cuckoo search algorithm,the proposed method is utilized for feature selection and parameters simultaneously optimizing and tested on UCI data sets and NSL-KDD data sets.The results show that the proposed CS-GWO algorithm has a better optimization effect in feature selection and simultaneous optimization.