Human-Like Rewards to Train a Reinforcement Learning Controller for Planar Arm Movement | IEEE Journals & Magazine | IEEE Xplore