thanks a lot for your sharing your insight. I do use RAVE and progressive widening or unpruning and the proximity to last move heuristic in the UCT tree by giving extra bias to those nearby locations. I guess they are not big enough or the code has bugs.
John On Tue, Jan 11, 2011 at 7:06 PM, Petr Baudis <[email protected]> wrote: > Hi! > > On Wed, Jan 12, 2011 at 01:04:47AM +0100, Petr Baudis wrote: > > The usual method to give some aim to 19x19 MCTS is giving positive > > bias to nodes that are very close by the Common Fate Graph metric. > > ...that is, very close to the last move. > > Petr "Pasky" Baudis > _______________________________________________ > Computer-go mailing list > [email protected] > http://dvandva.org/cgi-bin/mailman/listinfo/computer-go >
_______________________________________________ Computer-go mailing list [email protected] http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
