๐ฃ 2023 AI CUP Autumn Contest has begun!
The theme of this contest is "Let AI Be The King of Go"
Do you know that Go is the most difficult-to-proficient board game due to its variation? ๐
To help AI keep up with humans๐ตโ๐ซ,
Google DeepMind dived into "Alpha Go" in 2014 and finally defeated Se-dol Lee two years later.๐ฎ
Unlike Western chess and Chinese chess, Go does not employ brute-force or game tree search in conjunction with alpha-beta search to explore possible tree-like structures, as it is impractical. ๐ข
Soโ๏ธ,
AlphaGo instead leveraged a strategy network, a value network, and a Monte Carlo tree search (MCTS) to identify move positions.๐
Furthermore, to enhance AlphaGo's chess prowess ๐ช๐ป,
Google DeepMind employed 'self-play' to reinforce the strategy network, achieving the goal of reinforcement learning.๐ธ
Are you also interested in Google DeepMind's trained models' ability to enable computers to defeat human chess players?๐คฉ
The Go Strength Imitation and Go Style Recognition Competition, organized by the Ministry of Education, has quietly started on September 1st. ๐ฅ
๐ฃโจ Don't hesitate to sign up now! โจ๐ฃ
Registration-related URLs:
