tting feedback from him... it would be nice for
people in general to have Mr. Kim's take.
I'm reluctant to ask for more of Mr. Kim's time, as he was very
generous to play this exhibition match in the first place.
Peter Drake
http://w
f dropped in even
though his program SmartGo was not entered.
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
Thanks, and thanks for participating!
Peter Drake
http://www.lclark.edu/~drake/
On Aug 9, 2008, at 4:40 AM, terry mcintyre wrote:
Peter, many thanks to you and all the other participants in the
Computer Go Tournament! It was fun watching and participating!
Here's hoping we have a
Does anyone know how to contact the primary developers of GNU Go? I
haven't had any response from [EMAIL PROTECTED] or [EMAIL PROTECTED]
(This is in regards to their medal and certificate from the
competition at the US Go Congress.)
Peter Drake
http://www.lclark.edu/~
rtant areas of the board" this summer, and they tend
not to help much on 9x9.
I hope to work on 13x13 in the near future, so I hope the 13x13 CGOS
will still be up (and there will be some competition).
Peter Drake
http://www.lclark.
There is this:
http://webdisk.lclark.edu/drake/publications/GAMEON-07-drake.pdf
The idea of gradual acceleration is intriguing, although I shudder to
think of setting the parameters for such a process.
Peter Drake
http://www.lclark.edu/~drake/
On Sep 5, 2008, at 10:43 AM, Michael
ry, thus losing a net three
points). If you try to pass as well, I stubbornly insist that the
stone is alive, thus restarting the game.
What prevents this sort of abuse? Is this one of those cases where the
tournament director has to adjudicate?
(This is not a problem under Chinese or AGA
moves played in the "resumed" game are merely virtual.
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
I agree -- the AGA rules are quite clear. Note that the British Go
Association has recently adopted the same rules.
Peter Drake
http://www.lclark.edu/~drake/
On Sep 16, 2008, at 8:12 AM, David Fotland wrote:
Finally, a plug for American rules: American rules are the same as
chinese
gs on their own.
I would probably simply use AGA rules, but just about all English
introductory books (e.g., "Learn to Play Go" by Janice Kim and Jeong
Soo-huyn) use the Japanese rules.
Peter Drake
http://www.lclark.edu/~drake/
On Sep 16, 2008, at 7:25 PM, Ross Werner wrote:
Als
ause you don't want to fill in your own territory), and then I
pass. The game has ended again, and we still have a dispute.
Peter Drake
http://www.lclark.edu/~drake/
On Sep 18, 2008, at 7:41 AM, Jeff Nowakowski wrote:
On Wed, 2008-09-17 at 21:39 -0700, Ross Werner wrote:
And, of course
I understand this method, I just don't see where the (translated)
Japanese rules explain such a method.
Peter Drake
http://www.lclark.edu/~drake/
On Sep 18, 2008, at 9:15 AM, Ben Shoemaker wrote:
- Original Message
From: Peter Drake <[EMAIL PROTECTED]>
I really c
instinct is to lay down a
terse and elegant set of rules and then deal with the consequences of
those rules, but perhaps that is a bad thing when teaching.
Peter Drake
http://www.lclark.edu/~drake/
On Sep 18, 2008, at 10:43 AM, Jeff Nowakowski wrote:
I think you should teach both area
focuses on a specific area of the game.
Any other suggestions for books?
Peter Drake
http://www.lclark.edu/~drake/
On Sep 18, 2008, at 12:15 PM, Jeff Nowakowski wrote:
Ok, then play some 9x9 games with area scoring rules as Dave Devos
suggested. I was making the same suggestion. Don
the opponent to stay on it, then random playouts
would evaluate that position as lost, even if the forced sequence
would
make it a win.
It's true, this is a problem; raw Monte Carlo fares poorly at reading
ladders.
Peter Drake
http://www.lclark.edu/~drake/
___
The article is in Chinese, so I have no idea if there's anything of
interest.
Peter Drake
http://www.lclark.edu/~drake/
On Sep 30, 2008, at 7:12 AM, John Fan wrote:
http://post.weiqi.tom.com/BA000A6427.html in Chinese.
A quick summary on the games -- He very rarely played 9x9 game
(Sorry, I meant to forward that to a Chinese-speaking colleague, not
to re-send it to the list.)
Peter Drake
http://www.lclark.edu/~drake/
On Sep 30, 2008, at 8:47 AM, Peter Drake wrote:
The article is in Chinese, so I have no idea if there's anything of
interest.
Peter Drake
Here's a start:
http://hal.inria.fr/docs/00/28/78/67/PDF/icin08.pdf
Gelly et al, The Parallelization of Monte-Carlo Planning
Peter Drake
http://www.lclark.edu/~drake/
On Oct 2, 2008, at 2:39 PM, Michael Markefka wrote:
Now, for the technical matter: Could somebody please point me
929]C[mundungus [5k?\]: hi
zj [7k\]: hi
]
;B[cd]BL[1795.411]
;W[ed]WL[1790.725]
;B[dc]BL[1793.565]
;W[ec]WL[1788.904]
What are those items marked BL and WL? Time left?
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
comput
We've got some decent results (or at least interesting pictures) by
looking at the correlation between controlling a particular point and
winning the game.
Peter Drake
http://www.lclark.edu/~drake/
On Oct 3, 2008, at 7:47 AM, Claus Reinke wrote:
Do these make sense? And are there
I use JUnit unit tests.
Peter Drake
http://www.lclark.edu/~drake/
On Oct 22, 2008, at 6:09 AM, Mark Boon wrote:
I'm using unit-tests also, although somehow for Go programming not
nearly as much as I usually do.
And I use CruiseControl. It monitors changes in my repository,
build
those concerned about such things, there is some profanity.)
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
You might also try Orego. It's based around a Java re-implementation
of Libego.
Of course, we'll likely rewrite the core routines this month. If
you're not in a hurry (and want to work in Java), you might wait for
that.
Peter Drake
http://www.lclark.edu/~drake/
On May 3
Is anyone else here going to the International Joint Conference on
Artificial Intelligence, July 11-17, Pasadena, California (USA)?
http://www.ijcai.org/
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-go@computer
d a single stone to a chain? Do you walk
the enlarged chain counting liberties (and checking for duplicates
within the temporary liberty list), create a new one-stone chain and
then perform merging, or something else?
Thanks,
Peter Drake
http://www.lclark.
adjust the liberty counts
accordingly.
At most 3 checks are required. I only have to walk a full chain
when a move
merges two or more chains. I hope this is clear :)
Crystal clear -- thanks!
Peter Drake
http://www.lclark.edu/~drake
Next question: what about captures? Do you have to re-walk the
neighboring chains when a capture occurs?
Peter Drake
http://www.lclark.edu/~drake/
On May 31, 2009, at 9:27 PM, David Fotland wrote:
1) yes. I maintain liberty counts during MC playouts.
2) Something else. I remove one
paper was this?
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
Due to other commitments, I will NOT be attending the US Go Congress
in Fairfax, Virginia, August 1-9. As a result, if there is to be any
computer Go event there (e.g., computer-computer tournament, computer-
professional match), someone else will have to organize it.
Peter Drake
http
I'm for keeping the number of pools small, to keep their sizes large.
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
The Canadian overtime is new to me.
1) Can a program that simply never runs out of basic time safely
ignore this?
2) Is there something special that has to go in the config file? I
currently have this line:
rules.time=18:00
Thanks,
Peter Drake
http://www.lclark.edu/~drake
://webdisk.lclark.edu/drake/publications/BetaDistribution.pdf
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
On May 7, David Silver wrote:
However, if you can wait a few weeks I will be publishing a clearer
(I hope!) explanation of how to combine UCT and RAVE in my PhD thesis.
Did this happen? We're about to try implementing RAVE, so a clear
explanation would be a wonderful thing.
Peter
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.
Confused,
Peter Drake
http://www.lclark.edu/~drake/
Ah. I had always thought of "patterns" as translation-invariant.
Peter Drake
http://www.lclark.edu/~drake/
On Jun 22, 2009, at 11:45 AM, Magnus Persson wrote:
Probably 1x1 patterns implies that different priorities are assigned
to the absolute position of empty moves. AMAF can be
playing around with.
So, at the risk of sounding pedantic, these patterns aren't REALLY 1x1
-- they take into account other information, such as the number of
liberties a stone has. (Is this a correct interpretation?)
Peter Drake
http://w
work when using RAVE/AMAF?
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
I believe we used a uniform random policy (only "don't play in your
own pseudoeyes").
The numbers probably won't be the same, but we've certainly replicated
the qualitative improvement with version 6.05 of Orego, available here:
https://webdisk.lclark.edu/drake
sing? That presumably doesn't get
counted as a "good future move" in AMAF, so how can passing gain wins
(or runs, for that matter) when it's the right move (e.g., when there
is a seki on the board)?
Peter Drake
http://www.lclark.edu/~drake/
_
the win rate among
RAVE moves.
My question is: what is the meaning of this line?
coefficient = 1 - rc / (rc + c + rc * c * b)
Why this formula?
Thanks for any help you can offer,
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailin
y cool)
concerns solely the question of which move to select when your
search is over. The conventional choice is to play the move that has
the most *trials*, and Pebbles version is to play the move that has
the most *wins*.
Peter Drake
http://www.lclark.edu/~drake/
__
" to get rid of the old
nodes? Unfortunately, I can't fit a second copy of the table in
memory...
Thanks,
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
in there, but the machine thrashes when I go from 128K
to 256K nodes.
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
rching when there are no reusable
nodes.
...because this improves the quality of information in the nodes you
were able to create. That makes sense.
Thanks,
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-go@c
mercial version
allocates 30K nodes per CPU core. The version in the world
championship had
much more, but the commercial version can't be that greedy for memory.
That's reassuring.
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go
Aha, got it!
Thanks.
Peter Drake
http://www.lclark.edu/~drake/
On Jul 5, 2009, at 10:17 PM, David Fotland wrote:
Between moves, I find the nodes that can be recycled and put them on
a free
list. If the free list is empty I do a very short search, then give
up.
David
-Original
st one would always limit
the search to a single chain, without looking at other chains in the
hash table, wouldn't it?
I wasn't aiming for that level of sophistication; just hoping to reuse
nodes that are no longer in the tree (dag) because they're al
only traversing the
part of the tree that is still relevant (probably a small fraction of
the nodes in use), then traversing the set of nodes linearly to
rebuild the free list. In other words, I would perform a mark-and-
sweep garbage collection.
Peter Drake
http://www.lclark.edu/~drake/
performing playouts, but they would
greatly speed up the mark phase of marking and sweeping.
Yes, I think this would be an improvement. Thanks!
Peter Drake
http://www.lclark.edu/~drake/
On Jul 6, 2009, at 8:15 PM, Michael Williams wrote:
I thought you had really heavy nodes? Like 1k bytes or
On Jul 6, 2009, at 10:09 PM, Peter Drake wrote:
I suppose I could also include first child / next sibling pointers.
I wouldn't actually use them when performing playouts, but they
would greatly speed up the mark phase of marking and sweeping.
Hmm. That would work for a tree, but this
Terry McIntyre, Martin Mueller, and I will be meeting in at the IJCAI
registration desk at the Pasadena (CA) convention center at 12:30 on
Tuesday, July 14.
If any other computer Go people are in the area, we'd love to have you
join us.
Peter Drake
http://www.lclark.edu/~
As in LibEGO, if you define the off-board points to be both black AND
white, finding captures requires fewer branches.
Peter Drake
http://www.lclark.edu/~drake/
On Jul 13, 2009, at 8:48 AM, David Fotland wrote:
I use one dimensional arrays for speed (to avoid a multiply by 21).
Old Many
At long last, I ran across a situation where positional and situational superko differ. Unfortunately, Orego uses the former (I'll fix this presently) while KGS uses the latter.In this game, Orego (white) tries to move at S3 at the end. This would make the board look exactly as it did THREE turns a
Preallocate in advance. Allocating memory on the fly is expensive.
Peter Drake
http://www.lclark.edu/~drake/
On Jul 14, 2009, at 8:06 AM, Carter Cheng wrote:
This is something I have been curious about since I am somewhat new
to writing code in languages which require explicit memory
icular node may
have data generated from positions that are technically different
due to GHI issues.
We only check for superko right before actually making a move (i.e.,
not in the tree or in playouts).
Peter Drake
http://www.lclark.edu/~drake/
You might look in the genetic algorithm literature, where they have to
do this for fitness-proportional reproduction. A useful buzzword is
"roulette wheel".
Peter Drake
http://www.lclark.edu/~drake/
On Jul 15, 2009, at 4:06 PM, Mark Boon wrote:
When using patterns during the
I must be missing something. Isn't the obvious trick:
int r = random();
int i = 0;
while (r > weights[i]) {
r -= weights[i];
}
return i;
This way, you only have to generate one random number.
Peter Drake
http://www.lclark.edu/~drake/
On Jul 15, 2009, at 8:55 PM, Zach Wegn
On Jul 15, 2009, at 11:19 PM, Steve Kroon wrote:
This code should work with "r -= weights[i++]" in the loop body, and
comes down to a linear search through
You're right -- I forgot to increment i.
2009/7/16 Peter Drake
I must be missing something. Isn't the
I've recently been getting an odd distorted buzzing with every sound
played by CGoban3, the KGS client. This doesn't happen with other
applications, so I don't think it's a hardware or driver problem.
Has anyone else encountered this?
Peter Drake
http://w
I was looking at this game that Orego played against a human on KGS recently:
Orego-Zanarkand.sgf
Description: Binary data
I note that Orego's dead stones are marked as dead, but Zanarkand's are not. Does KgsGtp defer to the human when there are disputes about dead stones? Is that the most likely
active?
Thanks,
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-go@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/
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailma
made after ALL nodes in
the search, a sort of global RAVE table? It would be noisier still,
but it would fill with useful data very quickly, no?
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-go@computer-go.org
http
hrieb Peter Drake:
Is anyone also using a global table of moves made after ALL nodes
in the search, a sort of global RAVE table? It would be noisier
still, but it would fill with useful data very quickly, no?
Peter Drake
http://www.lclark.edu/~
listinfo/computer-go/
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-go@computer-go.org
http:/
rko during playouts,
just simple ko.
Do other programs handle this issue differently?
Peter Drake
http://www.lclark.edu/~drake/
On Aug 5, 2009, at 2:20 PM, Brian Sheppard wrote:
I analyzed the following position as a win for O, but there are
two or three kos involved (A1/A2, H1/G1, and the
ry (again, assuming self-atari
processing) so the position is almost bounded.
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
Actually, it's even worse than this: following LibEGO, my playouts
allow (multi-stone) suicide!
I may fix this before this weekend's KGS tournament.
(Speaking of which, where are all the contestants?)
Peter Drake
http://www.lclark.edu/~drake/
On Aug 5, 2009, at 8:10 PM, Peter D
rash, but it
does throw away such playouts.
To fix this, I plan to always check for superko violations. Is this
what others are doing, or is there another way out?
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
compute
part:
"otherwise it looks for the moves capturing stones on the Go board"
Does this really mean traversing the entire board looking for
captures? Doing so seems to create a catastrophic speed hit.
Peter Drake
http://www.lclark.edu/~drake/
___
y the size of the chain.
- From those moves that capture, choose the one that captures the most
stones.
Is there a better way?
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
e the speed cost. I'll run a larger experiment
tonight...
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
I realize that most games in the KGS computer tournaments end in
resignation, but just in case:
Is there a phase for removing dead stones (as in games with humans),
or are dead stones supposed to be captured (as in MC playouts)?
Peter Drake
http://www.lclark.edu/~drake
Just the info I needed -- thanks!
Politeness to human opponents aside, it sounds like a safe policy is
to never pass if there are dead enemy stones on the board.
Peter Drake
http://www.lclark.edu/~drake/
On Sep 9, 2009, at 4:44 AM, Jason House wrote:
If you're asking about stone cl
Orego is currently using the MoGo policy (escape, local patterns,
capture, random). Including these as priors helps a little, but just
including them in the playouts helps a lot, even with time (rather
than # of playouts) fixed.
Peter Drake
http://www.lclark.edu/~drake/
On Sep 12, 2009
n this list, David Silver gives a
newer formula:
http://computer-go.org/pipermail/computer-go/2009-May/018251.html
Was this ever published? (Orego is using this newer formula, and it
appears to work well.)
Peter Drake
http://www.lclark.edu/~drake/
__
We added (MoGo's original) patterns and RAVE at about the same time.
Both helped a great deal, and using both was best of all.
Peter Drake
http://www.lclark.edu/~drake/
On Sep 15, 2009, at 5:28 AM, Olivier Teytaud wrote:
Hi David, Thanks for these information.
Your patterns ar
U Go, 9x9 games,
giving Orego 1 second per move.
Of course, your mileage may vary.
Peter Drake
http://www.lclark.edu/~drake/
pastedGraphic.pdf
Description: Adobe PDF document
___
computer-go mailing list
computer-go@computer-go.org
http://www.
3.7.11, at the default level 10.
Specifically:
gnugo --mode gtp --quiet --chinese-rules --capture-all-dead --
positional-superko
Peter Drake
http://www.lclark.edu/~drake/
On Sep 16, 2009, at 9:19 AM, David Fotland wrote:
What version of gnugo, at what level?
-Original Message
lly defer to humans when there
are disputes? Isn't there supposed to be a cleanup phase? Would it be
the same in a rated game?
Orego-jika.sgf
Description: Binary data
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing li
Got it -- thanks!
Peter Drake
http://www.lclark.edu/~drake/
On Sep 17, 2009, at 3:55 PM, Jason House wrote:
This comes up from time to time on this list. Rated games require
the human to accept what the bot says (but can undo to continue
play). In free games the bots must accept what the
By far the strongest, cleanest code yet. If you're looking to get
started in computer Go and prefer to work in Java, this is the code
for you!
See command-line-options.txt in the doc directory.
http://legacy.lclark.edu/~drake/Orego.html
Enjoy!
Peter Drake
http://www.lclark.edu/~
I don't have a spare machine for that, but if someone else wants to
run one, I have no objection.
Peter Drake
http://www.lclark.edu/~drake/
On Sep 19, 2009, at 5:01 AM, Folkert van Heusden wrote:
Great!
Will you let it play on cgos?
On Fri, Sep 18, 2009 at 10:57:12AM -0700, Peter
ll of the white stones at the top dead, and
Orego disagreed. A cleanup phase was entered, and I believe housebot
crashed during the cleanup phase.
I don't know why that left the game alive with no clock ticking...
Peter Drake
http://www.lclark.edu/~drake/
On Sep 21, 2009
k A, white B, and white wins, I should (weakly) consider B as
a response to any black first move.
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
Yup, I tried something like that, too, with a similar lack of luck.
Peter Drake
http://www.lclark.edu/~drake/
On Sep 25, 2009, at 6:39 AM, Brian Sheppard wrote:
I have another way to fail to improve on RAVE. :-)
I tested a method that gives higher weight to recent RAVE data. The
method
Yes. I believe Fuego does this. See also Helmbold and Parker-Wood,
"All-Moves-As-First Heuristics in Monte-Carlo Go":
(Does anyone have a URL for this one? I can't seem to find it online,
but I have a paper copy in front of me.)
Peter Drake
http://www.lclark.edu/~drake/
ng information" problem I mentioned earlier.
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
Orego will enter.
Peter Drake
http://www.lclark.edu/~drake/
On Sep 28, 2009, at 12:45 PM, Nick Wedd wrote:
The October 2009 KGS computer Go tournament will be this Sunday,
October 4th, in the Asian evening, European morning and afternoon,
and American night, starting at 08:00 UTC/GMT (09
I've got Orego's account authorized to play ranked games. Is there
something I have to put in the kgsGtp config file to make it offer
ranked games?
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-g
Nope, I've got mode=custom. Perhaps it's automatch.rank=2k.
Peter Drake
http://www.lclark.edu/~drake/
On Oct 1, 2009, at 11:06 PM, David Fotland wrote:
This is my ini file. I think mode-custom does it, and gameNotes is
displayed with the game offer
engine=file-name-here -thread
Problem solved: you have to personally log into that account, open
user's information, and check the "rank" box.
Peter Drake
http://www.lclark.edu/~drake/
On Oct 2, 2009, at 3:32 AM, Eric Dunham wrote:
I have just used my admin power to un-disable Orego's rank, but it
o before adding
anything to the tree, but not verifying that the "best" move was
actually legal.
That's fixed now. Sorry about the inconvenience.
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-go@compu
Incidentally, if a new version of kgsGtp appears, the one feature I
desperately want is a way to tell kgsGtp to disconnect after the
current game. As it is, I have to either wait for the end of the game
or kill my program in the middle of someone's game.
Peter Drake
http://www.lclar
erritorial positions, leaving the hard-to-use outer influence
for its opponents.
Good advice!
We have, of course, linked goals: we want to make the program
stronger, but we also want to discover techniques that might be
relevant to problems other than Go; thus the focus on automatically
extrac
Many of us have concluded that, with RAVE, there is no need for a UCT
exploration term:
http://computer-go.org/pipermail/computer-go/2009-June/018773.html
Is there a published source on this result that I could cite?
Thanks,
Peter Drake
http://www.lclark.edu/~drake
I'm actually looking for something weaker than what Olivier has
offered: a published report of the empirical finding that (for some
programs, at least) an exploration coefficient of zero works best.
Peter Drake
http://www.lclark.edu/~drake/
On Nov 9, 2009, at 10:15 AM, Olivier Te
0]; it is now 0."
Peter Drake
http://www.lclark.edu/~drake/
On Nov 9, 2009, at 10:31 AM, Petr Baudis wrote:
On Mon, Nov 09, 2009 at 10:18:25AM -0800, Peter Drake wrote:
I'm actually looking for something weaker than what Olivier has
offered: a published report of the empirical findin
le of articles (not mine) up there. Others,
please follow suit!
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
ourse, it's somewhat slower with heavy playouts, but not THAT bad!)
If I only run playouts (no tree), I can get those up to 35 and 8 kpps
with a single thread.
Peter Drake
http://www.lclark.edu/~drake/
___
computer-go mailing list
compute
101 - 200 of 310 matches
Mail list logo
