Re: [computer-go] Purpose of Neural-Nets?

[Erik van der Werf <E.vanderWerf@xxxxxxxxxxxxx>]

Hi Chrilly,

chrilly wrote:
> In Go neural nets seem to be an inportant concept.
> Can the neural-net experts explain me the purpose? Is it primarly:
> a) To make the strongest Go-programm possible.
Yes, until I find something better.

> b) To learn network-learning.
No! For that I would strongly recommend a simpler domain than Go.

> c) To write a paper about network-learning.
Partially, but writing papers was never my primary goal, it's just 
something you have to do when you're working in a university.

> d) Its just fun to play around with neural-networks.
Sure, but many other techniques are fun also. The main thing is that I 
like learning, instead of having to hand code all the knowledge.


> In case of a). One could first learn the weights and then hardcode the
> weight-function to speed up the evaluation. 
Do you mean implement in hardware or just stop the learning?


Alternatively one could have
> 2 versions of a program. The slow learner and the fast
> tournament-player. Is this done, or is a neural-net programm always a
> neural-net program?
So far I never used learning during tournament games, and the code for 
the supervised learning was never even in the tournament program.


Hope this answers your questions.

Best,
Erik

_______________________________________________
computer-go mailing list
computer-go@xxxxxxxxxxxxxxxxx
http://computer-go.org/mailman/listinfo/computer-go