Attack rules: an adversarial approach to generate attacks for Industrial Control Systems using machine learning

Adversarial learning is used to test the robustness of machine learning algorithms under attack and create attacks that deceive theanomaly detection methods in Industrial Control System (ICS).Given that security assessment of an ICS demands that an exhaustive set of possible attack patterns is studied, in this work, wepropose an association rule mining-based attack generation technique. The technique has been implemented using data from aSecure Water Treatment plant. The proposed technique was ableto generate more than 300,000 attack patterns constituting a vastmajority of new attack vectors which were not seen before. Automatically generated attacks improve our understanding of thepotential attacks and enable the design of robust attack detectiontechniques.