Published Paper Details:

Abstract: Machine learning has been adopted in a wide range of domains where it shows its superiority over other algorithms. These methods can also be integrated in cyber detection systems with the goal of supporting replacing the first level of security analysts. Although the complete automation of detection and analysis is an still a distant goal, the efficiency of machine learning in cyber security must be evaluated with the due diligence. We present an analysis, addressed to security specialists, of machine learning techniques applied to the different types of cyber attacks. The goal is to assess the current maturity of these solutions and to identify their main limitations that prevent an immediate adoption of those machine learning cyber detection schemes. Our conclusions are based on an extensive review of the literature as well as on experiments performed on real enterprise systems and network traffic in different conditions.