Newsgroups: comp.ai.neural-nets
From: jimmy@ecowar.demon.co.uk (Jimmy Shadbolt)
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!news.sprintlink.net!demon!ecowar.demon.co.uk!jimmy
Subject: Re: Complexity analysis of NNs 
Distribution: world
References: <1994Oct27.115825.8801@dcs.warwick.ac.uk>
Organization: Econostat
Reply-To: jimmy@ecowar.demon.co.uk
X-Newsreader: Simple NEWS 1.90 (ka9q DIS 1.21)
Lines: 19
Date: Thu, 10 Nov 1994 10:25:06 +0000
Message-ID: <784463106snz@ecowar.demon.co.uk>
Sender: usenet@demon.co.uk

In article <1994Oct27.115825.8801@dcs.warwick.ac.uk> Martyn.Amos@dcs.warwick.ac.uk writes:

>Greetings.
>
>I didn't see this mentioned in the FAQ, so....
>I'm interested in complexity analysis of neural nets (eg. predicting average
> time
>taken to train a net to find an optimal solution). 
>
>If anyone has any references or pointers I'd be very grateful.
..

IMHO most results are related to the VC-dimension and essentially
discrete; does anyone know how to relate complexity to the "effective"
number of degrees of freedom of the net or similar functional measure
of approximation/representation power?

Drago

