Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: chess learning vs better search

Author: Vincent Diepeveen

Date: 19:42:20 03/06/06

Go up one level in this thread


On March 05, 2006 at 09:19:08, walt irvin wrote:

>On March 04, 2006 at 11:52:07, Charles Roberson wrote:
>
>>On March 04, 2006 at 10:52:52, walt irvin wrote:
>>
>>>On March 04, 2006 at 09:56:10, Charles Roberson wrote:
>>>
>>>>
>>
>>>
>>>that may be true but im sure that 50,000 or 60,000 games is not that hard to
>>>do,,,,besides if u used a strong program instead of that random one i talked
>>>about you could store positions up until say the score of + or - 3.5 should also
>>>eliminate alot of positions,,,,,
>>
>>   Maybe its more effort than you think.
>>   First you want the best play possible which implies long time controls.
>>   I guess that is 1.5 hours per side to play the whole game which averages
>>   to 1.5 minutes per move.
>>   Well that would be 180,000 hours for 60,000 games.
>>   That is 20.5 years.
>>
>>    That looks like a long time, so what about dropping the quality of the
>>   search. Lets drop the time per move to 9 seconds. That's a 10x drop.
>>   So, you are still at 2.5 years to complete.
>>
>>    The math seems to be working against you on this.
>
>ok i agree the math is working against me on that idea,,,,,what about getting a
>huge database of already played games and use that to learn with ???? that
>should not take 20 years ???? im sure there are many many many computer games
>out there and for that matter gm games  that u could use to help your program
>learn faster ,,,,
>
>walter irvin

Look there is 2 approaches.

You have a row of 3 people.

The first one is a chessgrandmaster. The second one is a student who's still
learning (your learner program) and who tells to the third person (your
resulting chessprogram) what to play. Every time your third person wants to know
something, he asks the student to ask the grandmaster what to play and why.

So what you are trying to do is to design 3 programs. The risk you run is that
your third person might simply not be learning things very well as he asks the
wrong questions. Additional the real big problem is you'll have to still create
that chessgrandmaster to judge positions.

What i'm doing is just creating that chessgrandmaster. That makes the task a lot
easier, than also making all kind of learners and protocols between those
programs.







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.