Petr Baudis wrote:
On Wed, Dec 16, 2009 at 08:57:24PM +0100, Rémi Coulom wrote:
The problem is that I am not using the same set of features for
biasing the tree, and for playouts. Playouts only use fast light
features. The tree part uses slow complex features. In particular, I
use patterns of radius 3 and 4 in the tree, and only radius 3 in the
playouts. When 3x3 patterns are learnt together with r=4 patterns,
they get different gammas.
Interesting, your paper said that you are using patterns up to r=10,
did you find out that anything larger than r=4 is irrelevant in
practice?
I have also tried r=4 up to 10 in Erica, with considering each point's
liberty.
In my experiments, r=5 and 6 help a lot. And, larger patterns help mainly
for opening playing,
such as playing at the corner in the first move (by far Erica has no joseki
database completely).
Aja
_______________________________________________
computer-go mailing list
[email protected]
http://www.computer-go.org/mailman/listinfo/computer-go/
