Intrusion Detection Based on Key Feature Selection using Binary GWO

Today's computing arc network based computing. Intrusion security is one of the major challenges of such computing facilities that have to deal with every time without compromising the system performance. In practical, no such intrusion detection system is implemented that can guarantee hundred percent true detection of intrusion and threats. In this paper, we have proposed a method of network intrusion detection system using key feature selection based on binary grey wolf optimization (GWO) and neural network classifier. The proposed IDS can be installed on any strategic point of the network. By eliminating the insignificant features from dataset using GWO the size of dataset can be reduced hence reduction in training time of the classifier and storage for dataset. The simulation experiments with NSL-KDD dataset show the improved accuracy of proposed intrusion detection method with reduced feature set.