Adaptive state aggregation for reinforcement learning | IEEE Conference Publication | IEEE Xplore