Constructing continuous action space from basis functions for fast and stable reinforcement learning | IEEE Conference Publication | IEEE Xplore