computer-go: Machine Learning and Go

[Thore Graepel <guru@xxxxxxxxxxxxxxx>]

Hello everyone,

we have just completed the first stage of our project "A machine
learning approach to the game of go" and have summarised the results in a
little paper.  You can download the paper from

http://stat.cs.tu-berlin.de/~guru/publications.html

It is called "Go, SVM, Go" and we submitted it to NIPS2000.

I thought you might be interested in this because we use a new approach of
converting a graph-based representation into feature vectors that enable us
to learn some aspects of good and bad "shape". These methods are then used to
learn good Tsume-Go moves and to play on a 9x9 board after learning from game
records. I would be delighted to receive some comments and suggestions. Also, if
you have any questions, please feel free to ask.

Thore

-- 
Thore Graepel | Technical University Berlin
FB13 (CS) | FR6-9 (STAT) | Franklinstr. 28/29 | 10587 | Berlin | Germany 
+49 30 314 22577 (phone) | +49 30 314 25901 (fax)
guru@xxxxxxxxxxxxxxxxx | http://stat.cs.tu-berlin.de/~guru