A short description of Many Faces:
It's an MCTS full board searcher. For the tree it uses the UCT formula and
RAVE, with an exploration term, and an MFGO bias. It does progressive
unpruning up to a maximum of 30 moves per position. The unpruning decision
is based on rave and MFGO bias.
MFGO
Many of the tenukis are due to a simple problem that I haven't had time to
fix yet.
The old mfgo expert system suggested reasons/goals for moves, each with a
value. After a move was made and the position evaluated, each reason was
checked to see if the move actually achieved the goal. For
Only the last game had full use of better hardware, since it used a 64 bit
version with more total memory. If the time used is available we should see
that Many Faces used much more time on the last game.
From: computer-go-boun...@dvandva.org
[mailto:computer-go-boun...@dvandva.org] On Behalf
Hi David,
Many thanks for sharing this info with us. I am not a strong Go player
and have a question which I suspect many people on this list could answer:
On 30/12/2010 08:09, David Fotland wrote:
...
hand tuned. There are rules for not filling eyes, not making self Atari
(unless it is a
On 30/12/2010 08:19, David Fotland wrote:
...
The old mfgo expert system suggested reasons/goals for moves, each
with a value. After a move was made and the position evaluated, each
reason was checked to see if the move actually achieved the goal. For
example if the suggestion was “this
On Thu, Dec 30, 2010 at 10:26:49AM +, Robert Finking wrote:
Hi David,
Many thanks for sharing this info with us. I am not a strong Go
player and have a question which I suspect many people on this list
could answer:
On 30/12/2010 08:09, David Fotland wrote:
...
hand tuned. There
Many thanks
On 30/12/2010 10:34, Petr Baudis wrote:
On Thu, Dec 30, 2010 at 10:26:49AM +, Robert Finking wrote:
Hi David,
Many thanks for sharing this info with us. I am not a strong Go
player and have a question which I suspect many people on this list
could answer:
On 30/12/2010 08:09,
...
The two strongest programs since MoGo are Zen and Many Faces
You should not forget Erica by Aja Huang,
winning the gold medal on 19x19 in the Computer Olympiad 2010.
(Zen and Many Faces also participated, getting ranks 2 and 3.)
Ingo.
--
GMX DSL Doppel-Flat ab 19,99 Euro/mtl.! Jetzt auch
You should also give more credit to CrazyStone as an early strong program
that contributed many ideas, comparable to Mogo. Remi is Aja's advisor, so
Erica continues the CrazyStone thread.
David
-Original Message-
From: computer-go-boun...@dvandva.org [mailto:computer-go-
Hi!
On Thu, Dec 30, 2010 at 08:48:08AM +0100, Robert Jasiek wrote:
Congratulations also to all the theorists! Without their great
discoveries, programs would still be weak. Might somebody please
give an overview on the relevant theories and how they work?
http://pasky.or.cz/~pasky/go/ has
On 30.12.2010 20:40, Petr Baudis wrote:
if their insight to give large
weight to the center of the board is truthful or just something that
will gradually disappear as they attain more strength.
If they are all tuned to fit traditional databases / traditional expert
knowledge rules, then it
On 12/30/2010 01:58 PM, David Fotland wrote:
You should also give more credit to CrazyStone as an early strong program
that contributed many ideas, comparable to Mogo. Remi is Aja's advisor, so
Erica continues the CrazyStone thread.
I did mention CrazyStone, and the Sensei's page lists it
Hi Francois, Welcome
For reference I need about 100k playouts with
RAVE to get 50% winrate against GnuGo 3.8 L10.
Yes that's more or less expected. At least before the
big improvements (yet to come ;-)
In my case I do a lot of testing at 4x1 because
the games are around 15 seconds long
David Fotland: 076301cba853$99234ad0$cb69e0...@com:
You should also give more credit to CrazyStone as an early strong program
that contributed many ideas, comparable to Mogo. Remi is Aja's advisor, so
Erica continues the CrazyStone thread.
I'd like to add both Zen and Erica use the large
Perhaps the client viewer should have the ability to hide comments by
rank. Then anyone can be allowed to post, as long as they know that
not everyone will hear them.
On Thu, Dec 30, 2010 at 11:11 AM, Jeff Nowakowski j...@dilacero.org wrote:
On 12/29/2010 05:47 PM, Jacques Basaldúa wrote:
I
Sounds nice in theory, but in practice it will get super-confusing really
quick as different people will have different settings, so you will see many
half-conversations from people you can hear responding to people you can't.
On Thu, Dec 30, 2010 at 6:10 PM, Michael Williams
On 12/30/2010 06:10 PM, Michael Williams wrote:
Perhaps the client viewer should have the ability to hide comments by
rank. Then anyone can be allowed to post, as long as they know that
not everyone will hear them.
The problem with that is you end up with a bunch of disjointed chat
(people
Hi!
On Thu, Dec 30, 2010 at 07:08:27PM +0200, Francois van Niekerk wrote:
Some details about Oakfoam:
- UCT algo (surprise, surprise ;))
- RAVE
- Mogo 3x3 patterns
- Open Source under the BSD license
- Almost everything is adjustable at runtime using parameters
- Achieved a 1700 ELO
Hi Jeff,
When, do you think, did Mogo started dominating all the KGS computer events
and CGOS, and also was the first to extend that dominance from 9x9 to
19x19.?
In Computer Olympiad 2007, Steenvreter was gold medal on 9x9. At the final
match of 19x19, it's easily to see that Mogo and
Sorry, I might be wrong at RAVE. Maybe it should be: Sylvain proprosed the
idea of RAVE and David Silver proposed a new formula for RAVE.
Aja
-原始郵件-
From: Aja
Sent: Friday, December 31, 2010 9:20 AM
To: computer-go@dvandva.org
Subject: Re: [Computer-go] News on Tromp-Cook ?
Hi Jeff,
It's a good start. I haven't tested against gnugo on 9x9 for a long time so
I tried a short test his afternoon.
Many Faces, 1000 playouts per move, vs Gnugo 3.7.10, level 10.
Many Faces won 75.9% of 3669 games (+-1.4% confidence level).
It took me about 300 versions tested to get from 10% wins
This is certainly a good time to sit back and look at what got us here. The
following key ideas have been mentioned so far: UCB, MCTS, RAVE, Pattern and
Go knowledge during MC simulation.These ideas are all essential to a strong
MC based Go program.If we want to pick the most important idea that
I can say on a personal note that it was MoGo that first made me pay
real attention to MCTS. Even though it was maybe a transitional
program that early on wasn't strong enough yet to beat traditional
programs and later got quickly superseded by stronger MCTS variations.
And I think MoGo was the
Hi Fuming,
The idea of improving the quality of simulation is more earlier, than Mogo’s
paper, in the Appendix A of Remi Coulom’s CG2006 paper “Efficient Selectivity
and Backup Operators in Monte-Carlo Tree
Search”(http://remi.coulom.free.fr/CG2006/CG2006.pdf):
The choice of a more clever
-- Forwarded message --
From: Fuming Wang fuming...@gmail.com
Date: Fri, Dec 31, 2010 at 1:50 PM
Subject: Re: [Computer-go] News on Tromp-Cook ?
To: Aja ajahu...@gmail.com
Hi Aja,
Remi and S. Gelly's paper both come out in 2006,and I just checked that they
did not reference each
Hi Fuming,
Remi's CG2006 paper is published/released half an year earlier. CG2006 was on
2006/5/29-31 at Turin, Italy. Mogo's paper
(http://hal.inria.fr/docs/00/12/15/16/PDF/RR-6062.pdf) was released on 2006/11.
That's why Mogo's paper cited Remi's paper (please see reference [9] in this
On 31.12.2010 03:16, Fuming Wang wrote:
This is certainly a good time to sit back and look at what got us here. The
following key ideas have been mentioned so far: UCB, MCTS, RAVE, Pattern and
Go knowledge during MC simulation.These ideas are all essential to a strong
MC based Go program.
So
Monte Carlo go was around for a long time. See Bouzy's papers for example.
The UCT formula for balancing exploration and exploitation came from
research on the one-armed bandit problem, not related to go.
Mogo and Crazystone's contributions were to show that monte carlo go could
be competitive
28 matches
Mail list logo
