A Tractable Algorithm for Finite-Horizon Continuous Reinforcement Learning | IEEE Conference Publication | IEEE Xplore