Fuzzy Approximation for Convergent Model-Based Reinforcement Learning | IEEE Conference Publication | IEEE Xplore