[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

computer-go: To Mans Ullerstam Only



Sorry to bother other people. I tried to contact him but all the replies
got bounced back. Just hope he can get this reply through this mailing
list.

Song


----------------------------------
On Tue, 7 Jan 2003, Song Li wrote:

> Hi Mans,
>
> Nice to hear from you! Yes, I am going to work on it next semester, though
> I am not studying it right now. Hopefully I will begin working on it next
> week. Yes, I would be very happy to have you participate in it if you are
> interested, and basically, I agree all your suggestion about the GNU Go,
> that's just my plan.
>
> If you like, I may send you a copy of my proposal about the project I will
> work on.
>
> Best,
>
> Song
>
> On Sun, 5 Jan 2003, [iso-8859-1] Måns Ullerstam wrote:
>
> > Hi Song,
> > That sounds interesting. You said you were going to work on it next
> > semester. I guess that means you are studying right now. Can you tell me
> > more about the project, and especially how you see me taking part of it.
> >
> > I can strongly recommend you using Gnugo as the basis for any Go
> > research. There are many reasons for this:
> > - You get started with a framework with user interfaces, interfaces to
> > play over GTP or GMP or NNGS, etc
> > - You get a fairly good computer go player that you can compare your own
> > work with (this means that you can use Gnugo as the benchmark for your
> > own program.
> > - Gnugo can ofcourse be the performance measure of any algorithm, which
> > means that instead of self play, you can play against Gnugo to decide
> > good or bad Go players (for GA)
> > - Any progress you make, will actually matter, since you can send your
> > better program back to the Gnugo project, making the program better
> >
> > Best regards / Mans
> >
> > > -----Original Message-----
> > > From: Song Li [mailto:sli@xxxxxxxxxxxxxxxxx]
> > > Sent: den 4 januari 2003 19:26
> > > To: Måns Ullerstam
> > > Subject: Re: computer-go: Master Thesis about Go in Japan
> > >
> > >
> > > Hi Mans,
> > >
> > > I have written a proposal about computer Go but I have not
> > > got time to do it.  I am going to work on it the coming
> > > semester. It's also involved in the reinforcement learning
> > > and GA. If you like, we may talk about it in detail. Happy new year!
> > >
> > > Song
> > >
> > > On Sat, 4 Jan 2003, [iso-8859-1] Måns Ullerstam wrote:
> > >
> > > > I am interested in writing my Master Thesis about Go. My
> > > goal is to go
> > > > to Japan next fall and work with my Master Thesis and study
> > > Go at the
> > > > same time.
> > > >
> > > > My focus areas are Artificial Intelligence, Reinforcement
> > > Learning and
> > > > Genetic Algorithms. I have worked in the Gnugo project and
> > > the idea is
> > > > to use Gnugo as the basis for my work and try different
> > > interesting AI
> > > > approaches to enhance the level of play.
> > > >
> > > > I have heard that there is some research going on in Japan
> > > with Go and
> > > > hopefully someone on this list can help me finding a project or
> > > > research group interested in my skills.
> > > >
> > > > Thanks / Mans Ullerstam, Sweden
> > > >
> > >
> > >
> >
> >
>
>