Actor-critic reinforcement learning for tracking control in robotics | IEEE Conference Publication | IEEE Xplore