A Heuristic Reinforcement Learning Based on State Backtracking Method | IEEE Conference Publication | IEEE Xplore