[Computer-go] Experiment?

terry mcintyre terrymcintyre at yahoo.com
Thu Mar 18 12:00:46 PDT 2010



 From: Olivier Teytaud <teytaud at lri.fr>




>>do you think those are program-specific or a general problem with UCT (or more likely the way simulations are run)?
>>Can you post a few specific cases?
>

> Big easy semeais. Very clearly defined, so that they are trivial for humans, and very big, so that there are many possible
>  permutations of liberties. One can clearly understand on this kind of situations why UCT-like or MCTS-like algorithms 
> will not solve them.

Yes, I have observed cases of where an otherwise highly-skilled UCT/MCTS program will fail, given large well-defined (to humans) semeai. 

Is there any way to do higher-level analysis of such semeais, in the same manner as humans do, in order to correctly solve the problem with fewer resources? 


      
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://computer-go.org/pipermail/computer-go/attachments/20100318/1c50e2a3/attachment.html>


More information about the Computer-go mailing list