-Original Message-
From: [EMAIL PROTECTED]
To: computer-go@computer-go.org
Sent: Wed, 28 Feb 2007 11:09 PM
Subject: [computer-go] Re: Big board
This seems to have gotten stuck in various email delays, so I am resending.
Sorry in advance if you get 2, but I did not see it get through.
This seems to have gotten stuck in various email delays, so I am
resending.
Sorry in advance if you get 2, but I did not see it get through.
Cheers,
David
Begin forwarded message:
From: David Doshay <[EMAIL PROTECTED]>
Date: 27, February 2007 6:29:01 PM PST
To: computer-go
Subject: Re: [c
Don mentioned using PBIL in a previous Go program, and related work with
Compact Genetic Algorithms.
Evidently, CGAs are susceptible to massive parallelization - an FPGA can yield
a 1000:1 speedup over a workstation:
http://www.aifb.uni-karlsruhe.de/Lehrangebot/Winter2001-02/HardSoft/literature
One more thought:
It would be interesting to see the degree to which following a
proximity heuristic leads to the renormalizations looking cold.
Cheers,
David
On 28, Feb 2007, at 11:07 AM, David Doshay wrote:
I do agree with Alain that beginners mix too little and random
players too much.
> I'm actually kind of surprised at the dissimilarity between the
> normal and multinormal. I'd expect the multinormal to boil down to the
> normal, but it looks like the standard normal has additional terms.
the multivariate normal has the 1-d normal as a special case, but
instead of normalizi
I do agree with Alain that beginners mix too little and random
players too much.
I am most intrigued with the recent results from Dave Hillis,
where he shows what I have been calling a move towards a
"transition temperature" with a selected set of heuristics in
the playout. When he is willing to
Le mercredi 28 février 2007 16:49, Oliver Lewis a écrit :
> On 2/23/07, David Doshay <[EMAIL PROTECTED]> wrote:
>
> >
> > On 22, Feb 2007, at 9:03 PM, alain Baeckeroot wrote:
> > > ... I made very slow progress to formalize this ...
> > > But the whole stuff is rather coherent in my mind.
> >
> >
Hi,
I'm currently trying to install a bot in a KGS' room, and thanks to the
michna's tutorial (http://www.michna.com/kgsbot.htm), it works nice except
that it seems to be impossible to restrict the bot game to the assigned
room.
Is there a way to make the bot uncheck the "Open list?" box (custom
Actually, the example given (and the first element in the table) are
exactly what I stumbled upon. In the scope of MC, I think this style of
analysis is 100% correct for evaluation of leaf nodes. I guess time
will tell if assuming a conjugate distribution with prior
hyperparameters is a good
On 2/23/07, David Doshay <[EMAIL PROTECTED]> wrote:
On 22, Feb 2007, at 9:03 PM, alain Baeckeroot wrote:
> ... I made very slow progress to formalize this ...
> But the whole stuff is rather coherent in my mind.
Then I envy you. I have been trying to bring what I know
about MC in physics toget
> Maybe other simple solutions exist,
you might want to check out those distributions that magically
have nice properties with respect to the bayesian integral.
they're called conjugate priors, and lots of distributions have
nice, easy to calculate conjugate priors.
there's a table here:
http:/
Based on my analysis, estimating a moves probability of winning by
taking the number of winning simulations (w) and dividing it by the
total number of simulations (n) is actually biased. I tried to break
this e-mail up into sections for easy digestion by the various people
who might read this
> I like your program and I actually tested it against yours during
> developement. But I didn't test Go169, what is that?
>
I changed the name from Go81 to Go169 at the time I switched from
PalmOS4 to PalmOS5. It also includes some 5x5 patterns which help with
boards larger than 9x9=81. (
On Wed, 2007-02-28 at 10:12 +0100, Heikki Levanto wrote:
> On Wed, Feb 28, 2007 at 09:28:37AM +0100, Magnus Persson wrote:
> > >I have never really understood the idea in all-in-first.
> >
> > Actually early in the game the order of random moves does not matter. The
> > Sequence AbCdE... have sam
On Wed, 2007-02-28 at 10:22 +0200, Tapani Raiko wrote:
> > "Ownership map" is a good term!
> >
> >
> Go81 (and Go169) also uses the ownership map (since 2002). In Palm
> handhelds, I can afford to do just two playouts, so the ownership map is
> much more informative than the first moves. I look
On Wed, 2007-02-28 at 09:01 +0100, Heikki Levanto wrote:
> > 2. I use a hybrid form of all-moves-as-first and others have
> > reported no improvement. The behavior I get is that it
> > plays much stronger at low simulations and in extensive
> > testing I could not find a high
On Wed, Feb 28, 2007 at 09:28:37AM +0100, Magnus Persson wrote:
> >I have never really understood the idea in all-in-first.
>
> Actually early in the game the order of random moves does not matter. The
> Sequence AbCdE... have same result if for example A and C switch order for
> CbAdE... If the
> "Ownership map" is a good term!
>
>
Go81 (and Go169) also uses the ownership map (since 2002). In Palm
handhelds, I can afford to do just two playouts, so the ownership map is
much more informative than the first moves. I look for large neutral
areas (especially moves that are close to both w
Quoting Heikki Levanto <[EMAIL PROTECTED]>:
I have never really understood the idea in all-in-first. That seems to
fly in the face of the common sense idea that the order of moves is
important. But perhaps I have got it wrong in my mind. I will have to
study more.
Actually early in the game th
On Tue, Feb 27, 2007 at 05:13:12PM -0500, Don Dailey wrote:
> This is very similar to what AnchorMan on CGOS does. At the end of
> each random simulation I keep the same statistcs on each point of
> the board and I use it to improve the move selection algorithm. I
> call this special board an "o
20 matches
Mail list logo
