Abstract: This article proposes a novel discounted inverse reinforcement learning (DIRL) algorithm for linear quadratic (LQ) control of unknown continuous-time (CT) systems with partially observable ...