Deception in Supervisory Control

The use of deceptive strategies is important for an agent that attempts not to reveal his intentions in adversarial environment. We consider a setting, which supervisor provides reference policy and expects follow the perform task. may instead different achieve model environment behavior with Markov decision process, represent tasks reachability specifications, study synthesis optimal policies such agents. also prevent supervisor's task high probability. show has convex optimization problem formulation, while requires solving nonconvex problem. NP-hard.