Sure, that would be a place to start. I also did some testing with just the number of pseudo-liberties at the position. That was pretty easy to code up. And I did have some limited success using 3x3 patterns-just not enough to justify the nuisance of carrying it along in my code.
There are quite a lot of choices involved. Testing them looks like the bottleneck to me. - Dave Hillis -----Original Message----- From: Michael Williams <michaelwilliam...@gmail.com> To: computer-go <computer-go@computer-go.org> Sent: Mon, Jun 22, 2009 4:28 pm Subject: Re: [computer-go] 1x1 patterns?! Why not start with the 5-vertex cross pattern? Going from 1x1 to 3x3 is a huge jump in complexity and debugability. With 5-vertex patterns, you can enumerate the patterns on paper (there are around 3^4 == 81 of them, ignoring symmetries) to sanity-check the details and see if the general idea works at all. There are too many to enumerate with 9-vertex 3x3 patterns (around 3^8 == 6561, ignoring symmetries).? ? dhillism...@netscape.net wrote:? > Yes. I think it's a good idea, but the devil is in the details. I've > become > pretty disenchanted with trying to use 3x3 or 5x5 patterns. > Currently, I > have about 300 1x1 patterns (I call them context codes) > that I'm playing > around with.? > > You can also do the same for RAVE without needing any more memory. You > > > only adjust the RAVE values, at a node, after filtering out moves, in > the > > playouts, that don't match the context/pattern for that position at > that > > node.? > > - Dave Hillis? > > > -----Original Message-----? > From: Michael Williams <michaelwilliam...@gmail.com>? > To: computer-go <computer-go@computer-go.org>? > Sent: Mon, Jun 22, 2009 3:02 pm? > Subject: Re: [computer-go] 1x1 patterns?!? > > I had never considered using AMAF with larger pattern. That's an > > > interesting idea. Perhaps a 5-vertex cross-shaped pattern or a 3x3 > > > pattern. Has anyone tried this? > > Magnus Persson wrote: > > Probably 1x1 > > patterns implies that different priorities are assigned > to > the absolute > > position of empty moves. AMAF can be seen this way. > AMAF > learns > > statistics of 1x1 patterns if the move is played in the > playout > but > > ignores all information surrounding the move at the time > it is > played. > > Another example would be to have lower priorities for > the moves > at the > > first and second line. > > > -Magnus > > > Quoting Peter Drake > > <dr...@lclark.edu <mailto:dr...@lclark.edu>>: > > >> I've seen reference in > > some papers to 1x1 patterns. What does that > even > >> mean? A point is > > either black, white, or vacant, and it's illegal to > >> play there unless > > it's vacant. > > > _______________________________________________ > > > > computer-go mailing list > > computer...@computer-g o.org <mailto:computer-go@computer-go.org> > > http://www.computer-go.org/mailman/listinfo/computer-go/ > > > _______________________________________________ > computer-go mailing list > computer-go@computer-go.org <mailto:computer-go@computer-go.org> > http://www.computer-go.org/mailman/listinfo/computer-go/ > > ------------------------------------------------------------------------? > Save energy, paper and money -- *get the Green Toolbar > > <http://toolbar.aol.com/green/download.html?ncid=emlweusdown00000037>.*? > > > ------------------------------------------------------------------------? > > _______________________________________________? > computer-go mailing list? > computer...@computer-go.org? > http://www.computer-go.org/mailman/listinfo/computer-go/? ? _______________________________________________? computer-go mailing list? computer...@computer-go.org? http://www.computer-go.org/mailman/listinfo/computer-go/?
_______________________________________________ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
