An Othello evaluation function based on Temporal Difference Learning using probability of winning | IEEE Conference Publication | IEEE Xplore