Friday, May 01, 2009

Kasparov in a Machine

Can you teach a computer program to play in the style of, say, Kasparov or Kramnik? So wondered Mark Levene and Trevor Fenner of Birkbeck College, in London.

We describe a preliminary investigation into learning a Chess player's style from game records. The method is based on attempting to learn features of a player's individual evaluation function using the method of temporal differences, with the aid of a conventional Chess engine architecture. Some encouraging results were obtained in learning the styles of two recent Chess world champions, and we report on our attempt to use the learnt styles to discriminate between the players from game records by trying to detect who was playing white and who was playing black.

The study can be downloaded from here. Hat tip to the Technology Review.

No comments: