Re: [computer-go] : Topological Invariant?

["Weimin Xiao" <weiminxiao@xxxxxxxxxxx>]

My feeling is that the method may not compress any, or do a real
abstraction, at all from a full board view. Any change in colored groups
changes empty position, and then you have a different graph.

Weimin

----- Original Message ----- 
From: "Eric Baum" <ebaum@xxxxxxxxxxxxxxxxx>
To: "computer-go" <computer-go@xxxxxxxxxxxxxxxxx>
Sent: Thursday, February 12, 2004 8:49 PM
Subject: Re: [computer-go] Re: Topological Invariant?



It is true that there are a lot of slightly different graphs,
but I don't know of any better compression. This does get a lot
of compression. This compression
already is so big that it loses game playing information unless
you further annotate it. Yet you seem to be asking for a bigger
compression than this, which is simultaneously still sufficient
to specify game position. I don't believe it exists.

 > You linked empty positions as they are on real board. Thus, any
 > small changes on the empty positions will change the graph. Or,
 > most likely, a little change in a colored group may change the
 > links to the empty positions. These make me feel the representation
 > is not topological invariant.

 > Weimin

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