RRT-based maximum entropy inverse reinforcement learning for robust and efficient driving behavior prediction | IEEE Conference Publication | IEEE Xplore