Newsgroups: comp.ai.neural-nets
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!gatech!concert!sas!mozart.unx.sas.com!saswss
From: saswss@hotellng.unx.sas.com (Warren Sarle)
Subject: Re: He who knows what he does not know is wise
Originator: saswss@hotellng.unx.sas.com
Sender: news@unx.sas.com (Noter of Newsworthy Events)
Message-ID: <Cz0yJq.GFJ@unx.sas.com>
Date: Wed, 9 Nov 1994 23:57:26 GMT
References: <parkCyxFB0.5Ko@netcom.com> <Cyz0MF.Jx3@unx.sas.com> <x05XzZ-.predictor@delphi.com>
Nntp-Posting-Host: hotellng.unx.sas.com
Organization: SAS Institute Inc.
Lines: 16


In article <x05XzZ-.predictor@delphi.com>, Will Dwinnell <predictor@delphi.com> writes:
|> Possibly a second neural net might be constructed, which would be trained
|> to predict the level of error made by the first net, using the same
|> inputs?  In effect, it would chart the input parameter space in terms
|> of the degree to which it had been modeled well.

But you have no target values for the second net, and the training data
for the second net would have to cover the entire sample space to some
reasonable density; this would be impractical with more than a few
inputs.
-- 

Warren S. Sarle       SAS Institute Inc.   The opinions expressed here
saswss@unx.sas.com    SAS Campus Drive     are mine and not necessarily
(919) 677-8000        Cary, NC 27513, USA  those of SAS Institute.
