Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Research idea? re: weight optimization

Author: Uri Blass

Date: 02:41:02 12/29/01

Go up one level in this thread


On December 29, 2001 at 05:11:26, Tom Kerrigan wrote:
<snipped>
>Are there any glaring problems with this idea?

Possible problems:
1)There are a lot of positions when we do not know the best move and it is not
clear that GM's moves are better than other moves that chess programs suggest.

It is possible to  choose only moves that your program can find after a long
search but in this case you need to waste a lot of time to find big number of
positions

It is also possible to use only positions that humans are sure that the GM's
move is the only best move but in this case you are not going to have a big
number of positions

2)It is possible that after you change your evaluation based on position A and
change your evaluation based on position B your program is unable again to find
the right move in position A.

Uri



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.