Re: Fuseki and Joseki Database with Neural Network

[Jens Yllman <jens.yllman@xxxxxxxxxxxxxxxxxx>]

 I'm not an expert on ANN. But I'm not sure joseki is best solved with ANN.
Because my experience with ANN is that it is not exact. But ANN can "guess"
where it is not sure. So maybe ANN is good for choosing fuseki and maybe
choose joseki depending on the fuseki. But to store a joseki sequence I
think normal datamodels is better than ANN.

 I'm trying to get some sparetime to work on my GO program. This program
uses ANN to selelect moves. Later on I may have time to studdy special
goals in GO. And see if ANN is especially good at some parts of GO.

 Jens Yllman

At 00:09 1998-06-17 +0200, you wrote:
>Hi there!
>
>Has someone experience of making a Fuseki / Joseki database with a
>neural network?
>Does it makes sense?
>
>I'm thinking about to let the NN offers some possible moves. Each of
>this proposals have to be evaluated and the best is taken.
>Maybe only to ask the NN when the normal database is exhaust; or the
>complete database is in the NN.
>
>It's only an idea. But could it possible, that a NN would be able to
>find a pattern in the beginning of a game and is able to make really
>good moves?
>
>Any experience or ideas?
>
>	Juergen

------------------------------------------------------------
Jens Yllman                       http://www.uniweb.se/~jens