An Interdisciplinary Approach to Machine Learning for Critical Infrastructure Protection

Critical infrastructure protection faces increasing challenges, both in quality and in quantity. Most of the present security systems fully rely on automated mechanisms, which replace human operators, in order to perform computation intensive tasks and/or to work in extreme conditions. However, this solution presents some drawbacks with respect to the system performance. In order to provide effective measures against the pressure of new and sophisticated threats, an interdisciplinary approach, based on suitably coupling machine learning with human judgment, results as the right choice. In fact, this solution is particularly helpful for implementing efficient solutions capable of controlling critical scenarios and reacting effectively towards sophisticated threats. This paper discusses the proposed approach and demonstrates that this approach is the best choice for the effective protection of critical infrastructures.