This interesting -- it implies that the place to use the heuristics
IS in the tree rather than in the playouts.
Peter Drake
http://www.lclark.edu/~drake/
On May 24, 2007, at 8:50 AM, Chaslot G (MICC) wrote:
Dear all,
I did experiments on 19x19 Mango with 25000 simulations per move,
against GnuGo 3.6 level 0.
Without progressive unpruning, Mango wins 2.9% (250 games), and
with progressive unpruning, Mango wins 31% (400 games).
I proposed progressive unpruning in this paper: http://
www.cs.unimaas.nl/g.chaslot/papers/pmcts.pdf
Simultaneously, Remi Coulom proposed the same thing in this paper:
http://remi.coulom.free.fr/Amsterdam2007/ under the name
“progressive widening”.
Question for native English speakers: do you think this technique
is best described by “progressive unpruning” or “progressive
widening”?
Cheers,
Guillaume
_______________________________________________
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
_______________________________________________
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
