Date of Publication :18th April 2018

Abstract: This paper is based on a hybrid intrusion detection system by using integrating K-means clustering and J48 classification. Firstly, the features are selected using correlation based feature selection, so that the number of attributes participating in detection of attacks can only be taken into concern and then it reduces the dimensionality of the attributes using Principal Component and Analysis. This algorithm works on the NSL-KDD dataset which is an improved version of the previously used KDD CUP’99 Dataset. Then we apply K-Means clustering over the obtained attributes and lastly we apply J48 classification for its evaluation. The proposed work has been fulfilled with an increase in accuracy and decrease in False Positive Rate

Reference :