Modular Reinforcement Learning architectures for artificially intelligent agents in complex game environments | IEEE Conference Publication | IEEE Xplore