Newsgroups: comp.ai.neural-nets
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!usenet.eel.ufl.edu!usenet.cis.ufl.edu!caen!msunews!harbinger.cc.monash.edu.au!news.cs.su.oz.au!metro!news
From: root@pc-philby.ae.su.OZ.AU (root)
Subject: Re: A simple but daunting question
Message-ID: <D2v1Ho.974@ucc.su.OZ.AU>
Sender: news@ucc.su.OZ.AU
Nntp-Posting-Host: pc-philby.ae.su.oz.au
Reply-To: root@pc-philby.ae.su.OZ.AU
Organization: Aeronautical Engineering, University of Sydney
References: <D2tIrF.Kqz@news.cis.umn.edu>
Date: Mon, 23 Jan 1995 13:29:47 GMT
Lines: 24

Hi Ed,

Now I'm not really one to endorse nets for stock market prediction, but I think you are
after a recurrent network. That is that only way to get a real dynamic mapping. The dilemma you will now face is how to extract static data from a dynamic system. It is possible in certain situations, whereby you feed the net the same pattern over and over until the output stabilises - dodgy, but it does work on occasions

Hope it helps

Phil


/****************************************/

Phil Blythe
Department of Aeronautical Engineering
University of Sydney
Australia

email phil@pc-philby.ae.su.oz.au

/***************************************/

The ideas and opinions expressed here are my mother's...


