On Oct 21, 2010, at 1:53 AM, Petr Baudis wrote:
On Wed, Oct 20, 2010 at 04:24:43PM -0700, Christoph Birk wrote:
But once I use UCT the improvement of AMAF (not RAVE) is only
minimal.
I use the terms
AMAF: all move as first (after the simulations)
RAVE: use the AMAF values to seed the (UCT) tree search
I'm sorry, I don't understand what do you mean by AMAF in the UCT
context. Can you elaborate?
I agreed with others that it (AMAF) does not improve the strength after
a certain point with random playout or UCT.
I was _not_ talking about RAVE which uses the AMAF statistics
for seeding the UCT search.
Matthew: Yes, RAVE gives an improvement with UCT.
Christoph
_______________________________________________
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
