Neural networks and differential dynamic programming for reinforcement learning problems | IEEE Conference Publication | IEEE Xplore