Abstract

This paper proposes Cuckoo Search Optimization (CSO) with Support Vector Machine (SVM) for the intrusion detection system (IDS). This work covers modules including preprocessing, feature selection and classification. The pre-processing is carried out using min-maximum standardization to remove missing values and filter the redundancy characteristics from the specified NSL KDD cup data set. Preprocessing helps primarily to increase the accuracy of the description. Instead CSO is used to pick the most suitable and optimum functions. With CSO, the search efficiency is improved and then the analysis is carried out more effectively to classify the intrusions using the SVM algorithm. This classification algorithm is used to increase the accuracy of attack detection. The test results show that the CSO with SVM algorithm is more efficient than existing methods.