Search found 417 matches

by Rémi Coulom
Wed Jun 06, 2007 2:40 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: A Go algorithm for chess programmers to try ?
Replies: 14
Views: 8049

Re: A Go algorithm for chess programmers to try ?

Hi Gerd, A few questions related to the domain specific features. You somehow index the database of your learned values by a move - like indexing a history table? Each feature has a kind of let say persistant history table with former learned values. And if featuers of a move have in combination (we...
by Rémi Coulom
Wed Jun 06, 2007 7:37 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: A Go algorithm for chess programmers to try ?
Replies: 14
Views: 8049

Re: A Go algorithm for chess programmers to try ?

I'm having a blast reading papers covering UCT-MC and related papers on variations to Bandit search and MCMC (Markov Chain Monte Carlo). MvH Dan Andersson This has become a hot topic in computer games. Out of 23 papers in the Amsterdam workshop, 7 deal with MC search. I predict it will be the domin...
by Rémi Coulom
Sun Jun 03, 2007 2:18 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: A Go algorithm for chess programmers to try ?
Replies: 14
Views: 8049

Re: A Go algorithm for chess programmers to try ?

You are NOT going to implement in Le Fous Fous (the crazy bishop) something that is in advance going to lose for you. I really believe it could work. Not Monte-Carlo chess, of course. But I was under the impression that using domain knowledge to prune or reduce late moves is a good idea. I am not f...
by Rémi Coulom
Sat Jun 02, 2007 8:32 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: A Go algorithm for chess programmers to try ?
Replies: 14
Views: 8049

A Go algorithm for chess programmers to try ?

Hi, I have developed an efficient algorithm for move evaluation in my Go program that, I believe, might be interesting to you. It can be used to learn move ordering heuristics from a database of sample good moves. It can also provide a probability distribution over legal moves in a previously unseen...
by Rémi Coulom
Fri Jun 01, 2007 7:48 am
Forum: Computer Chess Club: General Topics
Topic: Open Letter & $100,000 challenge to FIDE
Replies: 85
Views: 37734

Re: Open Letter & $100,000 challenge to FIDE

Hi, I'm making a $100,000 challenge (together with an open letter) to FIDE, for a match between Rybka and the winner of FIDE's upcoming 'Ultimate Computer Chess Challenge 2007'. More info is here: http://www.rybkachess.com/docs/ChallengeToFIDE.htm If anybody could help me to spread this to other ch...
by Rémi Coulom
Thu May 31, 2007 6:43 am
Forum: Computer Chess Club: General Topics
Topic: Participants WCCC 2007
Replies: 108
Views: 48966

Re: Participants WCCC 2007

I have no problem with Toga as a GPL deriviate of Fruit which is obviously making some people happy. I have no problem with gridchess as a research tool/subject I do have a problem with either of them entering in the WCCC. Richard. Hi, I share your point of view, but what can the organizers say if ...
by Rémi Coulom
Sun Apr 29, 2007 9:21 pm
Forum: Computer Chess Club: General Topics
Topic: How come CEGT is not consistent in displaying data?
Replies: 1
Views: 1045

Re: How come CEGT is not consistent in displaying data?

Hi, Uncertainty margins depend on individual games results. In particular, draws are more efficient than win and losses in reducing uncertainty. Two draws against the same opponent reduce uncertainty much more than one win and one loss. At least, this is how it works in bayeselo. You'll notice that ...