Re: computer-go: Semeai graph

[Joshua Berryman <joshuab@xxxxxxxxxxxx>]

I've seen one other piece of work treating go as a graph, although Im sure
there are completely different ways of approaching the problem:

http://stat.cs.tu-berlin.de/~guru/publications.html,

    Thore Graepel, Mike Goutrie, Marco Krger, and Ralf Herbrich. Learning
on Graphs in the Game of Go. , 2001.
    submitted to ICANN 2001.

Although Im sure Thore could have told you that himself. Perhaps its still
bedtime in Berlin. Of course this feeds into a whole bundle of work on
graphs+learning generally, an interesting paper for me was:

http://citeseer.nj.nec.com/schadler97connectionist.html, I think there
might be other ways to go with these ideas than the route taken above.

Josh



On Thu, 30 Aug 2001, Floris Geerts wrote:

> When searching the web using the key words " graph ", " game", and "go", I
> got the following two talks given by Katsuhiko Nakamura of Tokyo Denki
> University, Saitama, Japan.
> 
>  * Seminarji Slovenskega drustva za umetno inteligenco (1999)
>    "Graph-Theoretic Analyses of Board Phases in the Game of Go"
> 
>  * ACC '99 9th Conference Advances in Computer Chess
>    "Analyzing Capturing Races and Seki Situations in the Game of GO by
> Semeai Graphs"
> 
> This last conference should have its proceedings published in "Advances in
> Computer Chess 9", but  this seems not to be published yet (and will
> probably be too expensive to buy anyway).
> 
> Can someone sketch the basic idea of this approach or point me to some
> literature?
> 
> Thank in advance,
> 
> Floris Geerts
>