[Computer-go] Question about Zone Mode in MoGo?

[Petr Baudis]

  hi!

On Tue, Jan 11, 2011 at 05:10:57PM -0500, Go Fast wrote:
> Watching StoneGrid aimlessly playing on the large board, I am thinking of
> implementing the Zone mode mentioned in the MoGo's original paper on 19x19
> board. I am thinking of applying the zone mode in the tree part only, i.e.
> giving higher weight to the zone where the last move of the node belongs to
> when evaluating the final position of each simulation.
> 
> Has anyone tried the zone mode other than the MoGo team? Is it successful?

  I asked about this in October 2009 and the reply was that noone is
really using this, not even Mogo since long ago.

  The usual method to give some aim to 19x19 MCTS is giving positive
bias to nodes that are very close by the Common Fate Graph metric.
I.e. Pachi gives something like 40 won playouts to nodes in CFGd=1,
and 20 won playouts to nodes in CFGd=2,3. When I introduced this, it
helped a great deal on 19x19.

  Unless you are doing something clever, you also probably really want
RAVE or progressive unpruning on 19x19 to focus the tree reasonably.

-- 
				Petr "Pasky" Baudis
Computer science education cannot make an expert programmer any more
than studying brushes and pigment can make an expert painter. --esr