Multi-agent cooperation by reinforcement learning with teammate modeling and reward allotment | IEEE Conference Publication | IEEE Xplore