A Hybrid Genetic and Back-Propagation based for Network Intrusion Detection System

Abstract

The network and internet has become the most prominent growth of several organisations in which the major concerns that faced till today is the growth of cyber threats. The most important problem faced by network services is the increase of malicious attacks. Several researchers have analysed and proposed different techniques to overcome this problem and still needs various improvements. In this paper, a Hybrid Genetic based Back-Propagation Artificial Neural Network (GA-BPANN) method is proposed to improve the performance of the detection rate of malicious attacks in NIDS. The KDD Cup99 dataset is used for analysing the proposed method in terms of different attacks namely DDoS, Probe, and Normal. The experimental results show that proposed model obtains a high performance rate when compared with other existing techniques which is feasible to protect the system from various threats.