Reinforcement learning method for continuous state space based on dynamic neural network | IEEE Conference Publication | IEEE Xplore