Re: [computer-go] Cellular automata

[David Stern <dhs26@xxxxxxxxx>]

People following the CA thread might be interested in the following
paper (submitted to the NIPS 2004 conference):  

http://www.inference.phy.cam.ac.uk/dhs26/publications/nips2004.pdf

It describes a simple probabilisic model for making territory
predictions.  The model is structured to reflect the topology of the
board and you could think about it as propagating information around the
board itself in a recurrant manner similar to the way a CA might do.  

Note that this represents a simple first step into a program of similar
research.  This work was done with Thore Graepel and David MacKay and we
would be interested in any comments people may have.

Personally I feel that recurrant models (reflecting the grid topology of
the Go board) seem very suitable for Go and mirror nicely the way that 
stones exert and transmit influence.

Best Regards,

David Stern

On Thu, 2004-07-22 at 19:44, Peter Drake wrote:
> Has anyone tried cellular automata as an approach to Go?
> 
> Peter Drake
> Assistant Professor of Computer Science
> Lewis & Clark College
> http://www.lclark.edu/~drake/
> 
> _______________________________________________
> computer-go mailing list
> computer-go@xxxxxxxxxxxxxxxxx
> http://www.computer-go.org/mailman/listinfo/computer-go/

_______________________________________________
computer-go mailing list
computer-go@xxxxxxxxxxxxxxxxx
http://www.computer-go.org/mailman/listinfo/computer-go/