以前版本的AlphaGo最初接受了数千人的业余和专业游戏的训练,学习如何玩Go。
AlphaGo Zero跳过这一步,学习通过自行玩游戏,从完全随机的游戏开始。
在这样做的时候,它很快超过了人类的玩法水平,并将以前发布的冠军头衔的AlphaGo的100场比赛打败了。DeepMind
https://deepmind.com/blog/alphago-zero-learning-scratch/
This AI Taught Itself to Play Go and Beat the Reigning AI Champion – Motherboard
https://motherboard.vice.com/amp/en_us/article/8x8wy4/this-ai-taught-itself-to-play-go-and-beat-the-reigning-ai-champion
Nature : Nature Research
https://www.nature.com/nature/journal/v550/n7676/full/nature24270.html