Network Intrusion Detection Using Multi-Criteria PROAFTN Classification

Network intrusion is recognized as a chronic and recurring problem. Hacking techniques continually change and several countermeasure methods have been suggested in the literature including statistical and machine learning approaches. However, no single solution can be claimed as a rule of thumb for the wide spectrum of attacks. In this paper, a novel methodology is proposed for network intrusion detection based on the multicriteria PROAFTN classification. The algorithm is evaluated and compared on a publicly available and widely used dataset. The results in this paper show that the proposed algorithm is promising in detecting various types of intrusions with high classification accuracy.