Abstract: This paper presents a Q-learning approach to solving the finite-horizon optimal control problem for Boolean control networks (BCNs) under disturbances. We first introduce a dynamic ...