A Study of Reinforcement Learning in a New Multiagent Domain | IEEE Conference Publication | IEEE Xplore