Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Finding An Optimum By Random Route Changes

Author: David Blackman

Date: 18:26:30 11/28/98

Go up one level in this thread


On November 26, 1998 at 08:20:42, Graham Laight wrote:

[ in reply to Bas Hamstra's post about simulated annealing to find
   good solutions to the travelling salesman
 ]

>Unfortunately, I don't think the method would work for chess, because a "good"
>path in the search tree is not good enough - if your opponent is competitive,
>you really need the "best" route.

I wonder if this is true. What we can find now is the "best" path according
to a 10 ply search. What we might find is a "good" path according to a
much deeper search. 20 ply, 60 ply, any depth you want. My past experiments
with random tree methods were not a great success. But maybe i just wasn't
trying hard enough. I will definately try again some time.



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.