Feasibility of Machine Learning Techniques to Reduce False Alarm in Intrusion Detection?

The sophisticated recent advances in networking increased the dependability of humans in Network technology everyday life. Now-a-days the number of attacks on networks got increased drastically. With the development of large open networks, security threats have increased significantly in the past two decades. So mitigating those attacks is one of the significant interests of researchers in the network security. The goal of an intrusion detection system is to provide a wall of defence to confront the attacks of computer systems on internet. Machine Learning algorithms have been successfully applied to intrusion detection; however the true positive and false positive trade-offs is always a major challenge in the choice of the algorithms. This paper explores the wide variety of machine learning algorithms focusing on the feasibility of these algorithms for the purpose of reducing false positives.