Inverse Reinforcement Learning via Function Approximation for Clinical Motion Analysis | IEEE Conference Publication | IEEE Xplore