Bayesian Optimization in AlphaGo improves self-play win-rate from 50% to 66.5%arxiv.org3 points·teekno··0 commentsOpen articleSaveView on HN