Reinforcement learning to control a commander for capture the flag | IEEE Conference Publication | IEEE Xplore