Controlling physical agents through reactive logic programmingUsing background knowledge to speed reinforcement learning in physical agents
