David Silver wrote:
Hi everyone,
Please find attached my ICML paper with Gerry Tesauro on automatically
learning a simulation policy for Monte-Carlo Go. Our preliminary
results show a 200+ Elo improvement over previous approaches, although
our experiments were restricted to simple Monte-Carlo search with no
tree on small boards.
Very interesting, thanks.
If I understand correctly, your method makes your program 250 Elo points
stronger than my pattern-learning algorithm on 5x5 and 6x6, by just
learning better weights. That is impressive. I had stopped working on my
program for a very long time, but maybe I will go back to work and try
your algorithm.
Rémi
_______________________________________________
computer-go mailing list
[email protected]
http://www.computer-go.org/mailman/listinfo/computer-go/
