Newsgroups: comp.ai.neural-nets
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!uunet!in1.uu.net!cs.utexas.edu!howland.reston.ans.net!pipex!warwick!yama.mcc.ac.uk!cf-cm!C.M.Sully
From: C.M.Sully@cm.cf.ac.uk (Chris Sully)
Subject: introduction of random noise - refs?
Message-ID: <1995Apr6.171534.9633@cm.cf.ac.uk>
Sender: C.M.Sully@cm.cf.ac.uk (Chris Sully)
Nntp-Posting-Host: aquamarine.cm.cf.ac.uk
Organization: University of Wales College of Cardiff, Cardiff, WALES, UK.
Date: Thu, 6 Apr 1995 17:15:33 +0000
Lines: 28

Somebody has suggested to me that introduction of random noise into
training/ validation/ testing data might 

1. improve generalisation performance to unseen data and
2. allow determination of whether the network can perform any better on the
data set.

I haven't been able to find a great number of related references but am informed that this is an established technique.

Does anyone have any pointers to related literature for me?

Many thanks in advance.

Chris.


========================================================
Christopher Sully
Ph.D. Student (Neural Networks)  
Department of Computing Mathematics (Room M1.36), 
University of Wales College of Cardiff,   
PO Box 916, CARDIFF CF2 4YN, Wales, UK.   
E-Mail: C.M.Sully@cm.cf.ac.uk
WWW: http://www.cm.cf.ac.uk/User/C.M.Sully
Phone:  +44 (0)222 874000 x6070
Fax:	+44 (0)222 666182
Home:	+44 (0)222 484494
=========================================================
