An Efficient NIDS by using Hybrid Classifiers Decision Tree & Decision Rules

In the field of internet, network based application plays a vital role, where data transfers mostly in digital forms in various formats from source to destinations. In this digital exchange of information there are several possibilities of attacks and vulnerabilities. Intrusion detection systems are widely used to protect networks. An efficient detection of intrusion from network data set is a big problem which receives more attention from the various research communities. Various data mining classification techniques such as J-48 and SVM are widely used by researchers on various data sets such as KDD cup-99, NSL-KDD dataset. In this research paper we are presenting an efficient network intrusion detection system by using hybrid classifiers decision tree and decision rules for NSL-KDD dataset. An experimental study were performed by using weka-3-6 and MATLAB tool for existing J-48 and hybrid method and various performance improvement parameters such as precession, fmeasure, tprate and true positive rates are calculated. KeywordsNetwork based Intrusion detection system, KDD Cup-99, NSL-KDD, J-48, Hybrid decision tree