Author: Dann Corbit
Date: 15:23:55 07/25/00
Go up one level in this thread
On July 25, 2000 at 18:18:32, Tom Kerrigan wrote: >On July 25, 2000 at 17:38:21, Dann Corbit wrote: > >>On July 25, 2000 at 14:12:46, Tom Kerrigan wrote: >> >>>It would be nice to make a change to your evaluation function and get immediate, >>>accurate feedback. >>> >>>So my idea is to get a huge collection of positions of known value (i.e., "white >>>has a stronger position") and then see how well the known values correlate to >>>the evaluation function's scores. >>> >>>Does anybody have any ideas for getting a high-quality collection of such >>>positions? Or any comments on this approach in general? >> >>On my ftp site, project Apocalypse for CAP has an analysis of about 7000 EPD >>positions. You could take any subset from that set as a basis. There are about >>5000 or so which are definitely verified, so I suspect you could use them. >>However, there are also some where the bm does not agree with the pm or the am >>is the pm. > >Isn't the CAP data calculated via computer? So already there's some "noise" from >an evaluation function. Indeed it is. Consider also the approach I suggested in a follow-up to KD. I think that it is going to require a heuristic approach, since reasonably, we don't really know the answers to any of the positions. All of them are educated guesses.
This page took 0.01 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.