Update AlphaZero plays chess man

The developers of Google announced AlphaZero –
update Go DeepMind that plays chess person, according to
edition VistaNews.

 

To create used view machine learning with
which algorithm there is no training sample, i.e. the input data with the response. Originally
the program AlphaZero has been focused on the game of go, and the system was able to win
the strongest in this sport. Further, the developers have provided a set of
basic rules for classical and Japanese chess Shogi, where a new version of DeepMind
Go could show also very good results.

Analysts computer applications emphasize that
absolute victory against the man at AlphaZero no, and about a third of the games takes place
in a draw. However, this is a significant achievement for such programs, and
the developers hope that their algorithm will help not only in chess, but also to
to develop drugs, for example – when is a qualitative
update.

Share