Machine Learning Approach to IDS: A Comprehensive Review

Due to the very fast growth of computer networks, Internet emerges as an important tool to obtain the desired information. As the data transferred using networks is rapidly increasing, simultaneously the possibility of security threats have also increased. Hence, there is a need to constantly monitor the network traffic in order to secure private networks. Intrusion occurs when a set of actions trade off the confidentiality, integrity, or availability of a system. Intrusion detection systems (IDS) raise alerts if any unusual network traffic is detected and thus remains critical for network safety. Utilization of machine learning led methodologies in anomaly-based detection has gained popularity in recent years. In this paper, we conduct a thorough survey of studies based on the intrusion detection systems by using machine learning algorithms and present comparison dependent on dataset used, data reduction approaches, type of classifiers used and the outcome achieved by such different algorithms.