A New Learning Algorithm for the Maxq Hierarchical Reinforcement Learning Method | IEEE Conference Publication | IEEE Xplore