Comparative Analysis of Machine Learning based Intrusion Detection Systems

A critical component of network security is intrusion detection, which involves locating and stopping illegal access. Machine learning has shown potential as a technique for intrusion detection. Using the algorithms efficiently and overcoming the current drawbacks is the key to develop next generation models for intrusion detection system. In this paper, a comparison of various evaluation metrics from different papers is provided. The number of data used for training and name of dataset is listed as well. Analysis is done to provide researchers about the recently implemented algorithms and give a scenario of current challenges that are faced in IDS. In the end, this paper will be a guide for future directions in which researchers can work.