Author: Miguel A. Ballicora
Date: 14:37:41 01/18/01
Go up one level in this thread
On January 18, 2001 at 16:54:14, Tom King wrote: >On January 18, 2001 at 14:31:22, Miguel A. Ballicora wrote: > [snip] >>I am sure that this has been tried before (or something similar). >>Has been tried in this context? what do you think? I think that it is >>a good idea if no book is present. So far, I like it. >> > >this sounds very much like the idea published in 1990 by Tony Scherzer et al. >entitled "Learning in Bebe". It was one of the papers in "Computers, Chess and >Cognition". This is an excellent paper, published in an excellent book. I >recommend it thoroughly. Thanks, is that book still available? >>I tried another one that I did not like much so far (did not test it thoroughly, >>though): In the evaluation function, I added a random term that depended on the >>hash signature of that position. In that way, it was random but *reproducible*. >>(the seed, is determined at the beginning of the game from the time). >>However, the play was not very random or started to play crazy. Maybe >>the random numbers used were not good or I did not tuned properly the >>parameters. More to test. >Hmm. Not sure what this would add to the search. Just seems like a bit of noise >to me.. oops... I did not explain my intention. It was on the subject but not very clear. It has nothing to do with learning. Yes, that is the idea: to add a bit of noise to the search. In that way, it is guaranteed that it will start different each game. At least, it is relevant for the first 10 moves or so, then the noise can be disabled. Of course, if you have an opening book is not important. Regards, Miguel
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.