Re: [Computer-go] UEC Cup

[Petr Baudis]

On Sun, Mar 15, 2015 at 03:06:52PM +0900, Hiroshi Yamashita wrote:
> CrazyStone won.
> 
> 1. CrazyStone
> 2. DolBaram
> 3. Aya
> 4. Nomitan
> 5. Zen
> 6. Go Trend
> 
> Go Trend is a new Taiwan's program.
> Go Trend team has 7 people and they all visit here.

A few blurry photos I took:


http://pasky-jp.soup.io/post/556768169/UEC-Cup-2015-exhibition-game-was-between

http://pasky-jp.soup.io/post/556769031/The-prize-winners-Right-of-Remi-is
http://pasky-jp.soup.io/post/556769152/All-qualified-participants

I got the impression that O Meien is part of the GoTrend team.
Is that correct?  Does anyone know any details about GoTrend?
(Is it a university or commercial project?  Any interesting algorithms?)

Petr Baudis
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] Fwd: Teaching Deep Convolutional Neural Networks to Play Go

[Darren Cook]

> To be honest, what I really want is for it to self-learn,...

I wonder if even the world's most powerful AI (i.e. the human brain)
could self-learn go to, say, strong dan level? I.e. Give a boy genius a
go board, the rules, and two years, but don't give him any books, hints,
or the chance to play against anyone who has had access to books/teaching.

Darren

___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] UEC Cup

[Kahn Jonas]

A few blurry photos I took:


http://pasky-jp.soup.io/post/556768169/UEC-Cup-2015-exhibition-game-was-between

http://pasky-jp.soup.io/post/556769031/The-prize-winners-Right-of-Remi-is
http://pasky-jp.soup.io/post/556769152/All-qualified-participants


Thank you Petr,

So you were there, but did not have pachi play?
Jonas
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] UEC Cup

[Kahn Jonas]

UEC student won a exhibition game against CrazyStone without handicap.


Thank you for keeping us in touch, Hiroshi!
Sad for Zen, nice for Nomitan and nice to see Dolbaram in the finals. It
will be interesting to see it play a pro.

Do you know if there is a kifu for the game between CS and the student?

Thank you
Jonas
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] UEC Cup

[Lukas van de Wiel]

Thank you for the frequent updates and the pictures.

I am most curious to learn why it is that Zen, who usually ends up very
high in these competitions, has been defeated by two newcomers in the
field. What are Dolbaram and Nomitam doing differently?...

Cheers
Lukas

On Mon, Mar 16, 2015 at 9:58 AM, Kahn Jonas 
wrote:

> UEC student won a exhibition game against CrazyStone without handicap.
>>
>
> Thank you for keeping us in touch, Hiroshi!
> Sad for Zen, nice for Nomitan and nice to see Dolbaram in the finals. It
> will be interesting to see it play a pro.
>
> Do you know if there is a kifu for the game between CS and the student?
>
> Thank you
> Jonas
>
> ___
> Computer-go mailing list
> Computer-go@computer-go.org
> http://computer-go.org/mailman/listinfo/computer-go
>
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] Fwd: Teaching Deep Convolutional Neural Networks to Play Go

[Álvaro Begué]

The human brain is not the most powerful AI, because it fails the "A" test.

I suspect bootstrapping is not very hard. I have recently written a Spanish
checkers program starting with no knowledge and I got it to play top-human
level checkers within a few weeks.

You can build a database of games as you go. The important thing is that
the games don't have to be played perfectly: They just need to be
significantly better than your current model, so you can tweak the model to
learn from them. In the case of checkers the "model" is the evaluation
function, for which I used a small neural network. In the case of go the
model could be an evaluation function or it could be something that assigns
probabilities to moves. In either case, a database with hundreds of
thousands of games played by the current version of the program is very
useful to refine the model. A few iterations of (1. build database using
current model, 2. train model using database) should figure out how to play
the game well without any human knowledge.

Álvaro.



On Mon, Mar 16, 2015 at 4:52 AM, Darren Cook  wrote:

> > To be honest, what I really want is for it to self-learn,...
>
> I wonder if even the world's most powerful AI (i.e. the human brain)
> could self-learn go to, say, strong dan level? I.e. Give a boy genius a
> go board, the rules, and two years, but don't give him any books, hints,
> or the chance to play against anyone who has had access to books/teaching.
>
> Darren
>
> ___
> Computer-go mailing list
> Computer-go@computer-go.org
> http://computer-go.org/mailman/listinfo/computer-go
>
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] UEC Cup

[村松正和]

2015-03-16 17:47 GMT+09:00 Petr Baudis :

>
> I got the impression that O Meien is part of the GoTrend team.
> Is that correct?  Does anyone know any details about GoTrend?
> (Is it a university or commercial project?  Any interesting algorithms?)
>
Yes, he is. In fact, O Meien sensei and Shi-Jim Yen sensei organized
their team by a financial support of Trend Micro.
Their team includes two professional programmers, and they made a
strong program so quickly...

 Masakazu
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] UEC Cup

[Rémi Coulom]

Here is a partial game record I saved during the game. I am sorry I did not 
save the full game as sgf. It is on the nngs server of the UEC, so I might be 
able to get it tomorrow.

Crazy Stone invaded too deeply, and died.

Tomorrow, the handicap will be 4 stones for DolBaram, and 3 stones for Crazy 
Stone.

Rémi



crazystone-sakakibara.sgf
Description: Binary data


On 16 mars 2015, at 17:58, Kahn Jonas  wrote:

>> UEC student won a exhibition game against CrazyStone without handicap.
> 
> Thank you for keeping us in touch, Hiroshi!
> Sad for Zen, nice for Nomitan and nice to see Dolbaram in the finals. It
> will be interesting to see it play a pro.
> 
> Do you know if there is a kifu for the game between CS and the student?
> 
> Thank you
> Jonas
> ___
> Computer-go mailing list
> Computer-go@computer-go.org
> http://computer-go.org/mailman/listinfo/computer-go

___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] UEC Cup

[村松正和]

As for the game records in the UEC Cup, we are checking them manually
because there are some garbage in the
SGF files saved by the NNGS server. It will take several days to clear
them all to publish on our Web page. Sorry.

By the way, as a side information, Sakakibara kun is a 7d in Tygem.

 Best regards,

 Masakazu


2015-03-16 18:37 GMT+09:00 Rémi Coulom :
> Here is a partial game record I saved during the game. I am sorry I did not 
> save the full game as sgf. It is on the nngs server of the UEC, so I might be 
> able to get it tomorrow.
>
> Crazy Stone invaded too deeply, and died.
>
> Tomorrow, the handicap will be 4 stones for DolBaram, and 3 stones for Crazy 
> Stone.
>
> Rémi
>
>
>
>
> On 16 mars 2015, at 17:58, Kahn Jonas  wrote:
>
>>> UEC student won a exhibition game against CrazyStone without handicap.
>>
>> Thank you for keeping us in touch, Hiroshi!
>> Sad for Zen, nice for Nomitan and nice to see Dolbaram in the finals. It
>> will be interesting to see it play a pro.
>>
>> Do you know if there is a kifu for the game between CS and the student?
>>
>> Thank you
>> Jonas
>> ___
>> Computer-go mailing list
>> Computer-go@computer-go.org
>> http://computer-go.org/mailman/listinfo/computer-go
>
>
> ___
> Computer-go mailing list
> Computer-go@computer-go.org
> http://computer-go.org/mailman/listinfo/computer-go
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] Fwd: Teaching Deep Convolutional Neural Networks to Play Go

[hughperkins2]

> The important thing is that the games don't have to be played perfectly: They 
>just need to be significantly better than your current model, so you can tweak 
>the model to learn from them.

Thats an important incite. I hadnt thought of that. 

Maybe could combine with some concept of "forgetting", eg weight decay, so the 
net gradually unlearns some of the original, more naive, associations? > The 
important thing is that the games don't have to be played perfectly: They just 
need to be significantly better than your current model, so you can tweak the 
model to learn from them.

Thats an important incite. I hadnt thought of that. 

Maybe could combine with some concept of "forgetting", eg weight decay, so the 
net gradually unlearns some of the original, more naive, associations? could 
combine with some concept of "forgetting", eg weight decay, so the net 
gradually unlearns some of the original, more naive, associations? ___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] UEC Cup

[Hideki Kato]

Lukasz,

They are not newcomers.  See KGS Annual championship ranking, for 
example.  Nomitan and DolBaram have KGS rank 3d and 5d, resp. in blitz 
games.  
#DolBaram registered last UEC Cup but didn't come to the venue :D

The reason Zen lost against Nomitan was, as far as I've studied, just 
a hard luck.  A major evidence is that the losing moves played at the 
venue were repeated in very few probabilities at my site where almost 
all conditions are the same (only the connection from my laptop to the 
main computer of the clsuter is different, ie, via Internet or lan).

#So far, Zen often lost territory and built thickness in early stages 
and revesed in hte middle.  However, rivals are quickly getting stronger 
while top two are almost standing :-( these a few years, it's also 
getting harder to reverse.  Additionally, Nomitan plays slower (compared 
to others, such as Aya) but steadily, reversing is much harder and the 
lost territory (bottom left in attached sgf) is bigger than usuall.  
Very hard luck.

Hideki

Lukas van de Wiel: 
:
>Thank you for the frequent updates and the pictures.
>
>I am most curious to learn why it is that Zen, who usually ends up very
>high in these competitions, has been defeated by two newcomers in the
>field. What are Dolbaram and Nomitam doing differently?...
>
>Cheers
>Lukas
>
>On Mon, Mar 16, 2015 at 9:58 AM, Kahn Jonas 
>wrote:
>
>> UEC student won a exhibition game against CrazyStone without handicap.
>>>
>>
>> Thank you for keeping us in touch, Hiroshi!
>> Sad for Zen, nice for Nomitan and nice to see Dolbaram in the finals. It
>> will be interesting to see it play a pro.
>>
>> Do you know if there is a kifu for the game between CS and the student?
>>
>> Thank you
>> Jonas
>>
>> ___
>> Computer-go mailing list
>> Computer-go@computer-go.org
>> http://computer-go.org/mailman/listinfo/computer-go
>>
> inline file
>___
>Computer-go mailing list
>Computer-go@computer-go.org
>http://computer-go.org/mailman/listinfo/computer-go
-- 
Hideki Kato 

zen-nomitan.sgf
Description: x-go-sgf
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] Fwd: Teaching Deep Convolutional Neural Networks to Play Go

[Oliver Lewis]

I am also interested in unsupervised learning. When this first came up, I
found this paper: http://arxiv.org/pdf/1312.5602v1.pdf, where the authors
taught a deep convolutional network to play games from the Atari console.
They used what they call "Deep Reinforcement Learning", a variant of
Q-learning, and the methodology is described in detail.  There's even an
online demo of this technique for an easier problem here:
http://cs.stanford.edu/people/karpathy/convnetjs/demo/rldemo.html

It's impressive that the same network learned to play seven games with just
a win/lose signal.  It's also interesting that both these teams are in
different parts of Google. I assume they are aware of each other's work,
but maybe Aja can confirm.



On Mon, Mar 16, 2015 at 11:51 AM, hughperkins2 
wrote:

> > The important thing is that the games don't have to be played perfectly:
> They just need to be significantly better than your current model, so you
> can tweak the model to learn from them.
>
> Thats an important incite. I hadnt thought of that.
>
> Maybe could combine with some concept of "forgetting", eg weight decay, so
> the net gradually unlearns some of the original, more naive,
> associations? > The important thing is that the games don't have to be
> played perfectly: They just need to be significantly better than your
> current model, so you can tweak the model to learn from them.
>
> Thats an important incite. I hadnt thought of that.
>
> Maybe could combine with some concept of "forgetting", eg weight decay, so
> the net gradually unlearns some of the original, more naive, associations?
> could combine with some concept of "forgetting", eg weight decay, so the
> net gradually unlearns some of the original, more naive, associations?
>
> ___
> Computer-go mailing list
> Computer-go@computer-go.org
> http://computer-go.org/mailman/listinfo/computer-go
>
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] UEC Cup

[Lukas van de Wiel]

Dear Hikedi,

thank you for the insight. I stand corrected. Also thank you for the SGF
file.

Cheers
Lukas

On Mon, Mar 16, 2015 at 12:34 PM, Hideki Kato 
wrote:

> Lukasz,
>
> They are not newcomers.  See KGS Annual championship ranking, for
> example.  Nomitan and DolBaram have KGS rank 3d and 5d, resp. in blitz
> games.
> #DolBaram registered last UEC Cup but didn't come to the venue :D
>
> The reason Zen lost against Nomitan was, as far as I've studied, just
> a hard luck.  A major evidence is that the losing moves played at the
> venue were repeated in very few probabilities at my site where almost
> all conditions are the same (only the connection from my laptop to the
> main computer of the clsuter is different, ie, via Internet or lan).
>
> #So far, Zen often lost territory and built thickness in early stages
> and revesed in hte middle.  However, rivals are quickly getting stronger
> while top two are almost standing :-( these a few years, it's also
> getting harder to reverse.  Additionally, Nomitan plays slower (compared
> to others, such as Aya) but steadily, reversing is much harder and the
> lost territory (bottom left in attached sgf) is bigger than usuall.
> Very hard luck.
>
> Hideki
>
> Lukas van de Wiel:  jes9qasg1dgg4duct32ckswk+jvh6xj26yndbksb8...@mail.gmail.com>:
> >Thank you for the frequent updates and the pictures.
> >
> >I am most curious to learn why it is that Zen, who usually ends up very
> >high in these competitions, has been defeated by two newcomers in the
> >field. What are Dolbaram and Nomitam doing differently?...
> >
> >Cheers
> >Lukas
> >
> >On Mon, Mar 16, 2015 at 9:58 AM, Kahn Jonas 
> >wrote:
> >
> >> UEC student won a exhibition game against CrazyStone without handicap.
> >>>
> >>
> >> Thank you for keeping us in touch, Hiroshi!
> >> Sad for Zen, nice for Nomitan and nice to see Dolbaram in the finals. It
> >> will be interesting to see it play a pro.
> >>
> >> Do you know if there is a kifu for the game between CS and the student?
> >>
> >> Thank you
> >> Jonas
> >>
> >> ___
> >> Computer-go mailing list
> >> Computer-go@computer-go.org
> >> http://computer-go.org/mailman/listinfo/computer-go
> >>
> > inline file
> >___
> >Computer-go mailing list
> >Computer-go@computer-go.org
> >http://computer-go.org/mailman/listinfo/computer-go
> --
> Hideki Kato 
> ___
> Computer-go mailing list
> Computer-go@computer-go.org
> http://computer-go.org/mailman/listinfo/computer-go
>
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] UEC Cup

[Aja Huang]

Hi Hideki,

2015-03-16 11:34 GMT+00:00 Hideki Kato :
>
> #So far, Zen often lost territory and built thickness in early stages
> and revesed in hte middle.  However, rivals are quickly getting stronger
> while top two are almost standing :-( these a few years, it's also
> getting harder to reverse.  Additionally, Nomitan plays slower (compared
> to others, such as Aya) but steadily, reversing is much harder and the
> lost territory (bottom left in attached sgf) is bigger than usuall.
> Very hard luck.
>

Zen and Crazy Stone are impressive even just being standing at 6d. But yes
we all hope that the strongest Go programs can advance to 7d even 9d. I'm
glad to see DolBaram becomes strong and GoTrend is catching up quickly.

Aja
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] Fwd: Teaching Deep Convolutional Neural Networks to Play Go

[Aja Huang]

Hello Oliver,

2015-03-16 11:58 GMT+00:00 Oliver Lewis :
>
> It's impressive that the same network learned to play seven games with
> just a win/lose signal.  It's also interesting that both these teams are in
> different parts of Google. I assume they are aware of each other's work,
> but maybe Aja can confirm.
>

The authors are my colleagues at Google DeepMind as on the paper they list
DeepMind as their affiliation. Yes we are aware of each other's work.

Aja
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] UEC Cup

[Stefan Kaitschick]

What an unfortunate game against the human opponent. The trouble started
early with the tight low pincer at move 12.
The way black could press down on that move is what created the incentive
for CS to play the totally unreasonable invasion on the right a little
later.


Stefan


On Mon, Mar 16, 2015 at 2:37 AM, Rémi Coulom  wrote:

> Here is a partial game record I saved during the game. I am sorry I did
> not save the full game as sgf. It is on the nngs server of the UEC, so I
> might be able to get it tomorrow.
>
> Crazy Stone invaded too deeply, and died.
>
> Tomorrow, the handicap will be 4 stones for DolBaram, and 3 stones for
> Crazy Stone.
>
> Rémi
>
>
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] UEC Cup

[Kahn Jonas]

Here is a partial game record I saved during the game. I am sorry I did not 
save the full game as sgf. It is on the nngs server of the UEC, so I might be 
able to get it tomorrow.

Crazy Stone invaded too deeply, and died.


Thank you Rémi!

That was a very surprising game. I would have sworn that CS was black.
Since when does a MC bot play subway go? An effect of NN?


Tomorrow, the handicap will be 4 stones for DolBaram, and 3 stones for Crazy 
Stone.


Best of luck for the handicap game. I hope we see our first 3-stone
victory!

Et félicitations pour la coupe :)
Jonas
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] Teaching Deep Convolutional Neural Networks to Play Go

[Oliver Lewis]

Can you say anything about whether you think their approach to unsupervised
learning could be applied to networks similar to those you trained? Any
practical or theoretical constraints we should be aware of?

On Monday, 16 March 2015, Aja Huang  wrote:

> Hello Oliver,
>
> 2015-03-16 11:58 GMT+00:00 Oliver Lewis  >:
>>
>> It's impressive that the same network learned to play seven games with
>> just a win/lose signal.  It's also interesting that both these teams are in
>> different parts of Google. I assume they are aware of each other's work,
>> but maybe Aja can confirm.
>>
>
> The authors are my colleagues at Google DeepMind as on the paper they
> list DeepMind as their affiliation. Yes we are aware of each other's
> work.
>
> Aja
>
>
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] implementing simple ko

[Ray Tayek]

On 3/12/2015 9:49 AM, Hiroshi Yamashita wrote:

...

i need to implement a simple ko rule.


I wrote simple pure-MC and UCT sample.
http://www.yss-aya.com/puremc20150313e.zip
put() function handles a simple ko.


got it.

thanks

--
Honesty is a very expensive gift. So, don't expect it from cheap people - 
Warren Buffet
http://tayek.com/

___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] implementing simple ko

[Ray Tayek]

On 3/12/2015 6:16 AM, Robert Jasiek wrote:

On 12.03.2015 06:07, Ray Tayek wrote:

i need to implement a simple ko rule.


Supposing you want a strategically simple AND complete ko ruleset, you 
can choose from:


1) short cycle loss (2 or 3 successive plays, see below) + long cycle 
tie ko rule


2) basic ko rule + fixed ko rule (recreating a same position is 
allowed but then a different play must be made)


If you do not search for a complete ko ruleset but for a practically 
easily implemented and feasible:


3) basic ko rule + 3-play-loss (for the player playing to recreate the 
position)


Note that none of these are used by go servers or in the real world. ...


nice summary.

thanks

--
Honesty is a very expensive gift. So, don't expect it from cheap people - 
Warren Buffet
http://tayek.com/

___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] implementing simple ko

[uurtamo .]

You can have multiple kos on the board. Two is the most usual case that
doesn't trigger other rules. Easiest (logically, not practically, perhaps)
is to never repeat a board position ever (so called superko, I think),
which could be implemented as a full-board-position hash.

s.
On Mar 11, 2015 10:08 PM, "Ray Tayek"  wrote:

> i need to implement a simple ko rule.
>
> looks like you need to keep track of 2 points for each ko and who took it
> last. so it looks like a list of (point,point,who) that can change each
> turn?
>
> any pointers will be appreciated.
>
> thanks
>
> --
> Honesty is a very expensive gift. So, don't expect it from cheap people -
> Warren Buffet
> http://tayek.com/
>
> ___
> Computer-go mailing list
> Computer-go@computer-go.org
> http://computer-go.org/mailman/listinfo/computer-go
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] UEC Cup

[Petr Baudis]

On Mon, Mar 16, 2015 at 06:37:14PM +0900, Rémi Coulom wrote:
> Tomorrow, the handicap will be 4 stones for DolBaram, and 3 stones for Crazy 
> Stone.

DolBaram won the 4-handicap game against Cho Chikun 9p!

Petr Baudis
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go