Re: Neural Network Literature

Eric Ruud (ejruud@ucdavis.edu)
Fri, 11 Jun 1999 21:09:53 -0700



From: Darin Sunley <rsunley@escape.ca>
To: <extropians@extropy.com>
Sent: Friday, June 11, 1999 6:06 PM
Subject: Re: Neural Network Literature

>
>
> Anders Sandberg wrote:
>
> > "Eric Ruud" <ejruud@ucdavis.edu> writes:
> >
> > > I've recently become quite interested in neural networks. What should
I =
> > > be reading?
>
> Depends a lot on your background. Most of the actual research in NN's
becomes
> very difficult to read without a good solid grounding in both graph theory
and
> matrix arithmetic.

Sounds painful :) This might be a project better saved for another time.

> Does UCDavis have a CompSci department? Maybe email some professor and ask
what
> textbooks the 3rd or 4th year AI classes are using. Those'll have pretty
decent
> intros to NNs. At least, mine did.

They do, but for the most part all of my contact with the faculty has shown me that either a) They seem not to know what they're talking about or b) They have almost no teaching skills whatsoever.

The programming classes that I've taken here have been... less than comfortable.

> The major truth that we've discovered about NNs, though, is that the
particular
> arrangement of neurons into higher level structures (gain controls,
feedback
> layers) is WAY more important then simply trowing training data at a bunch
of
> neurons.
>
> Darin Sunley
> rsunley@escape.ca
>

Thanks for all the help! At least I have a general direction now.

-Eric

"I want specifics on the general idea
I wanna think what I should know
Want you to do me what to show
I wanna see movies of my dreams"

-Built to Spill