High performance computers have surpassed human in Chess – the most popular board game in Europe/America, and is challenging human in the game of GO – the most popular board game in East Asia. This article elaborates the internal mechanisms and algorithms used in state-of-art computers designed to playing board games, and briefly comments on its implications in the AI perspective. The content covers topics including game-tree complexity, EXPTIME-complete problems, static evaluation functions, bounded minimax search/Alpha-Beta pruning, and Monte Carlo tree search. This article is published in the Chinese version of the popular science magazine "NEWTON". Thanks so much to Chen Fang for commissioning the article.
|Book title||NEWTON - Science World|
|Journal||NEWTON - Science World|
|Publisher||NEWTON - Science World|