Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Deep Thought's tuning code and eval function!

Author: Ian Osgood

Date: 09:58:26 09/06/00

Go up one level in this thread


On September 05, 2000 at 16:08:28, Gian-Carlo Pascutto wrote:

>On September 05, 2000 at 13:10:52, Severi Salminen wrote:
>
>>Hi!
>>
>>I checked Tim Mann's web pages and here is what I found:
>>
>>http://www.research.digital.com/SRC/personal/mann/deepthought.html
>>
>>I don't know if this is an old story, but anyway...
>
>Thanks for sharing this!
>
>I've only been able to give it a brief look but it looks very
>interesting...
>
>--
>GCP

This is fascinating!  It didn't appear to me that the techniques mentioned were
tied to the Deep Thought/Hitech architecture.  It would probably take some work
to adapt the method to an eval function which contains lots of conditional
branching (as opposed to the straight linear combination of terms and
piece-square tables as used by Deep Thought).

Note: actual working source code is included.



This page took 0.03 seconds to execute

Last modified: Thu, 07 Jul 11 08:48:38 -0700

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