Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Killer heuristics in other engines?

Author: Antonio Dieguez

Date: 19:59:49 03/31/04

Go up one level in this thread


On March 31, 2004 at 17:30:34, Dieter Buerssner wrote:

>On March 31, 2004 at 12:32:39, Antonio Dieguez wrote:
>
>>aaaah, I use a web service to obtain a synchronyzed priority list and stuff.
>
>I am not sure, I understand.

I recommend to write:
"I am not sure if I understand"
or:
"I am not sure about my understanding"

Am sorry, but I am doing something that is supossed to be a QA, so am getting a
little used to that stuff..

> Are you using a priority queue, when you generate
>the moves to store them, and then have use log(number_of_moves) proportional
>time to store generated moves, and also to get them out of the queue? I tried
>something similar, and it did not buy much. Move sorting time is just not a
>major factor of the whole search. I even tried ideas, like trying to guess,
>wether this is a fail low node (where sorting will not help much) or a fail high
>node. The guess seems decent, but still, it was not really better than an easy
>inner loop of a selection search, to pick the moves (selection search will need
>time proportional to number of moves squred).
>
>>Last time amyan played with kasparov it did fine.
>
>Perhaps I look stupid, and misundertand any humor. Are you serious?
>
>Cheers,
>Dieter

:) man, you are good, I toke some minutes to figure out you are joking, but am
not even completely sure..

am sorry I don't elaborate more, am not very good. Anyway, in amyan I have many
things, except efficient algorithms, so you will see O(n*n) stuff more often
than O(log(n)) (especially in the neural pattern recognition engine.)

Cheers,
Antonio



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.