Reinforcement learning based on a statistical value function and its application to a board game | IEEE Conference Publication | IEEE Xplore