On Tue, 31 Mar 2009, Matthew Woodcraft wrote:
Jonas Kahn wrote:
You might be interested by this article, for a very complete and tested
answer. Plus the idea of grouping, but a good part of the effect seems
to me to be giving a heuristic pre-value to moves, which might be done more
efficiently
You might be interested by this article, for a very complete and tested
answer. Plus the idea of grouping, but a good part of the effect seems
to me to be giving a heuristic pre-value to moves, which might be done more
efficiently otherwise:
Although Tei and Aoba Professionals explained the match at the
front stage with a projection, the game was so complicated that I
couldn't see which is winning until near the end. Another semi-final
match, my Fudo Go vs Katsunari, also was shown on the screen but in a
small picture at upper right
Part of the problems stem from that playouts are weak, and more
specifically notably weaker than the program itself.
To begin with, a consequence is that most areas of the board are less
clear than they should to playouts. This entails, I think, a preference
for probable points against sure
Wasn't it today that Crazystone had a match against a professional
player? During the FIT2008 conference at Keio University?
Does anyone know the result and if the game is available somewhere?
Jonas
___
computer-go mailing list
Congratulations to Mogo team!
Twenty years from now, in ``a computer go history''
August 7th 2008: First victory of computer against pro with 9 handicap.
By the way, the surge in strength with the 800 processors with respect
to the quadcore (old) MogoBot, seemed relatively low, when comparing to
On Wed, Apr 02, 2008 at 02:13:45PM +0100, Jacques Basaldúa wrote:
Jonas Kahn wrote:
I guess you have checked that with your rules for getting probability
distributions out of gammas, the mean of the probability of your move 1
was that that you observed (about 40 %) ?
If I understand your
By contrast, you
should test (in the tree) a kind of move that is either good or average,
but not either average or bad, even if it's the same amount of
information. In the tree, you look for the best move. Near the root at
least; when going deeper and the evaluation being less precise,
So I believe a better approach is a heavy playout approach with NO
tree. Instead, rules would evolve based on knowledge learned from each
playout - rules that would eventually move uniformly random moves into
highly directed ones. All-moves-as-first teaches us that in the
general case
I think there was some confusion in Don's post on ``out of atari'' in
play-outs.
For one thing, I do not agree with the maximal information argument.
Testing ``out of atari'' moves is not good because they might be good,
or might be bad, but merely because they might be good. By contrast, you
Hi Jacques
No. for a reason I don't understand, I get something like:
Distribution fit expected 0.1 found 0.153164
Distribution fit expected 0.2 found 0.298602
Distribution fit expected 0.3 found 0.433074
Distribution fit expected 0.4 found 0.551575
Distribution fit expected 0.5 found
On Tue, Mar 11, 2008 at 09:05:01AM +0100, Magnus Persson wrote:
Quoting Don Dailey [EMAIL PROTECTED]:
When the child nodes are allocated, they are done all at once with
this code - where cc is the number of fully legal child nodes:
In valkyria3 I have supernodes that contains an array of
Typically, how many parameters do you have to tune ? Real or two-level ?
I guess I have 10 real valued and 10 binary ones. There are probably a lot
of stuff that are ahrd coded and could be parameterized.
Here I am also completely ignoring playouts that have hundreds of handtuned
On Mon, Mar 10, 2008 at 02:33:03AM -0400, Michael Williams wrote:
Jonas Kahn wrote:
out, kos can go on for long. I don't know what depth is attained in the
tree (by the way, I would really like to know), but I doubt it is that
MoGo displays the depth of the principle variation in the stderr
On Mon, Mar 10, 2008 at 01:03:02PM -0700, Christoph Birk wrote:
On Mon, 10 Mar 2008, Petr Baudis wrote:
MoGo displays the depth of the principle variation in the stderr stream.
I have been wondering, does that include _any_ nodes, or only these
above certain number of playouts? What is the
I think the general outline is that you pre-test groups first to see if
a self-atari move is interesting.It's worthy of additional
consideration if the stones it is touching have limited liberties and
the group you self-atari is relatively small.Then you could go on to
other tests
# One question: where _aya_ comes from or stands for? If my guess is
correct, you are confusing Hiroshi, author of Aya, and I, Hideki,
author of GGMC :). I'm sorry if I'm wrong.
I did. Sorry for the confusion. :(
Jonas
___
computer-go mailing
From my observaion, mc chooses good moves if and only if the winning
rate is near 50%. Once it gets loosing, it plays bad moves. Surely
it's an illusion but it helps to prevent them.
If it's more important to avoid being too pessimistic (ie low estimated
winning rates), there are two ways
I don't see that, but then again I am not a very strong player
myself. What I notice is that it plays very normal until it's
pretty obvious that it's losing, not just when it varies slightly from
50% but when it doesn't vary much from zero. However, it does play
more desperately once
There is much high-level data to be found within the MC runs, such as
whether a group is alive or not, etc.
Now, I don't know if it is easy to inject it back within the
simulations.
Another approach (not excluding the first one) would be to gather much
lower-level data.
It's especially sad that
But correct ko threats playing has nothing to do with the playout part :
Since it is a strategic concept that involves global understanting, It is
handled by the UCT tree part.
Yes and no.
Theoretically, that's the work of the UCT part. But, as Steve pointed
out, kos can go on for long. I
http://ewh.ieee.org/cmte/cis/mtsc/ieeecis/tutorial2007/Bruno_Bouzy_2007.pdf
Page 89, which kind of outcome. This method is better than the above
and similar to what Jonas seems to propose. The improvement is minor.
By looking at their proposal (45 * win + score), in contrast to mine,
there
delta_komi = 10^(K * (number_of_empty_points / 400 - 1)),
where K is 1 if winnig and is 2 if loosing. Also, if expected
winning rate is between 45% and 65%, Komi is unmodified.
There's one thing I don't like at all, there: you could get positive
evaluation when losing, and hence play
These ideas are all old,
I never said they were new. I wanted to give a mathematical argument on
them.
What would have been new would have been methods with filters applied on
the \hat{p}_i. However, though I am pretty sure I could make them more
efficient with little data, that's certainly not
The idea of using f(score) instead of sign(score) is interesting. Long
ago, I tried tanh(K*score) on 9x9 (that was before the 2006 Olympiad, so
it may be worth trying again), and I found that the higher K, the stronger
the program. Still, I believe that other f may be worth trying.
In
I experimented with something similar a while ago, using the
publicly available mogo and manipulating komi between moves.
If its win probability fell below a certain threshold (and the move
number wasn't too high), I told it to play on the assumption that it
would receive a few points more
The professional player who commented the game between Katsunari and
Crazy Stone thought that at the end of fuseki, Katsunari was ahead.
I wonder: even if it might not be optimal, does Crazy Stone play what is
best for him, that is, what he knows best how to use ?
I mean, if Crazy Stone played
You have basically 2 cases when losing. One case is that the program
really is busted and is in a dead lost position.The other case is
that the program THINKS it's lost but really is winning (or at least has
excellent chances.) In the first case, we are simply playing for a
Hi there
I am new here, but have read the list for a few monthes.
I am a mathematician, finishing my PhD on quantum statistics (that is
statistics on quantum objects, quantum information, etc.).
So do not expect me to write any code, but I could have suggestions for
heuristics in the choice of
29 matches
Mail list logo
