RE: [computer-go] Learning : was Chess programs versus go programs

["David Fotland" <fotland@xxxxxxxxxxxxxxx>]

I'm also guilty of top-posting and not snipping.  It's easier.

I have not considered every possibility for improvement.  I've worked on
computer go
for a long time, but not very consistently.  I probably have put a total of
about 3 years
full time into it.

I do use alpha-beta+evaluation.  The issue isn't a/b, but the high amount of
pruning required by the slow evaluation function.  The real debate should be
whether one can make a good enough evaluation function without using local
tactical search inside the evaluation.  That's what makes evaluation so
slow.  I believe local search is required since if tactical stability of
stones is not established, there is no way to estimate territory.

David

> -----Original Message-----
> From: computer-go-bounces@xxxxxxxxxxxxxxxxx 
> [mailto:computer-go-bounces@xxxxxxxxxxxxxxxxx] On Behalf Of A van Kessel
> Sent: Friday, December 10, 2004 8:30 AM
> I Hope I can. I will.
> David is a very clever man who worked on his program for 
> approx 20 years. (not constantly, I know ...) He has thought 
> an written extensively about it, and from what I recall from 
> it, he has considered every possibility for improvement. If 
> you read the discussions about alpha/beta+evaluationfunction 


_______________________________________________
computer-go mailing list
computer-go@xxxxxxxxxxxxxxxxx
http://www.computer-go.org/mailman/listinfo/computer-go/