On Jul 24, 2009, at 10:41 AM, Isaac Deutsch <i...@gmx.ch> wrote:
Hi Jason,
OK I see that this alters the result.
The "full pattern set" is just all 3x3 patterns, so there isn't a
lot of additional knowledge. Nevertheless, there is a downwards
tendency for all patterns. You can find the exact patterns
(triangle, connect and hane) and the average rating of all patterns
here:
http://pastie.org/557743
Thanks,
Isaac
An overall drift in the numbers might be nothing. Some pattern (sub)
sets can be multiplied by a constant value without affecting overall
prediction accuracy. Fixing one or more gamma values may fix your
drift issue. I think RemÃ's paper forced the average gamma to be 1
after each iteration. _______________________________________________
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
