Reinforcement learning with continuous vector output | IEEE Conference Publication | IEEE Xplore