Killer moves works well in alpha-beta search I guess because it found
by searching the PV then when alternative moves are played to the best
they are cut of one at a time with the killer move. Thus the killer
move is applied to almost the same position in every attempt.
In a playout the boardstate is arbitraty every time a "killer move" is
attempted. So you may try a 5*5 pattern and collect statistics for
them. The problem is that either this pattern is either to large or
too small to capture the necessary information you need. The solution
may be a smarter way of defining patterns and learn them dynamically.
-Magnus
Quoting Yamato <[email protected]>:
(2011/05/25 22:44), Stefan Kaitschick wrote:
As to the problem of smarter playouts, has something along the lines of
the "killer heuristic"(the most successful response sofar to a specific
move) , used in chess programming, been tried?
There's a lot more housekeeping than with RAVE and AMAF, but I can't
believe it hasn't been tried.
Maybe the effort could be reduced by only storing the
response-successrates of the 5*5 surrounding area, or something like that.
Anyways, I haven't even heard of a failed attempt, which is a little
strange to me.
I suppose that is called "Adaptive Playout".
Hendrik Baier reported LGRF heuristics and other lots of failed methods.
www.ke.tu-darmstadt.de/lehre/arbeiten/master/2010/Baier_Hendrik.pdf
--
Yamato
_______________________________________________
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
