Newsgroups: comp.ai.neural-nets
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!cam-news-feed3.bbnplanet.com!news.bbnplanet.com!cpk-news-hub1.bbnplanet.com!EU.net!usenet2.news.uk.psi.net!uknet!usenet1.news.uk.psi.net!uknet!uknet!newsfeed.ed.ac.uk!leeds.ac.uk!news
From: TRAMCMDC@leeds.ac.uk (M.C.M. DE CARVALHO)
Subject: Comparison NN x Logit
Message-ID: <56n1g5$7v4_001@leeds.ac.uk>
NNTP-Posting-Host: its_pc101.leeds.ac.uk
Organization: University of Leeds
Date: Sun, 17 Nov 1996 12:47:00 +0000 (GMT)
X-Newsreader: News Xpress Version 1.0 Beta #4
Lines: 21

Hi,

I would like to receive some comments on the subject NN 
comparison to Logit models.

I would like to know the answer of:

Is NN without hidden units similar to Logit?

Why adding hidden units to the NN does not improve the 
results, once it is said NN can approximate any nonlinear 
function?


Please, if you have any comments on these questions, send it 
to my e-mail address: mclecide@its.leeds.ac.uk.

Thanks,

Cleci. 

