通用强化学习算法AlphaZero

Thegameofchessisthelongest-studieddomaininthehistoryofartificialintelligence.Thestrongestprogramsarebasedonacombinationofsophisticatedsearchtechniques,domain-specificadaptations,andhandcraftedevaluationfunctionsthathavebeenrefinedbyhumanexpertsoverseveraldecades.Bycontrast,theAlphaGoZero