Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Iterated search vs. Internal iterative deepening

Author: Robert Hyatt

Date: 06:00:56 06/21/02

Go up one level in this thread


On June 21, 2002 at 05:13:36, Russell Reagan wrote:

>In a post from yesterday Bob Hyatt mentioned many areas in computer chess that
>have been written about and published in academic papers. He mentioned both
>"iterated search" and "internal iterative deepening". I had only heard the term
>iterative deepening (not internal iterative deepening, or iterated search).
>
>So what is the difference between:
>
>Iterative Deepening
>Internal Iterative Deepening
>Iterated Search
>
>Russell

Iterative search in internal iterative deepening have some common ground.

I think you understand iterated search so I will pass on that.

Internal iterative deepening is an approach to use when you are searching
and you don't have a good move to search at a critical node.  IE suppose you
are searching the _first_ move at the root, trying to find a PV, and at
ply=1 thru 6 you have a hash table move to try first, a known good move to
try early.  But at ply=7, you don't find one.  Here you can, rather than
relying on the usual things like captures, killers, etc, try a shallow
search to find the best move.  Start at 2 plies, then 3, and assuming you
are going to search to depth=14 here, you continue thru depth=6.  Now you
have a reasonable move to try here that is probably best, so search that
move first to 7 plies then continue with the others.

You don't do this everywhere, just at PV nodes...



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.