Safe reinforcement learning for probabilistic reachability and safety specifications: A Lyapunov-based approach

Open in new window