Newsgroups: comp.ai.neural-nets
Path: cantaloupe.srv.cs.cmu.edu!rochester!udel!news.mathworks.com!newsfeed.internetmci.com!EU.net!sun4nl!cs.vu.nl!news
From: agtoever@cs.vu.nl (Oever ten AG)
Subject: Re: ANN architecture for Stock Market Forecasting
Nntp-Posting-Host: kits.cs.vu.nl
References: <4f1997$8c6@ctylnk.cityu.edu.hk>
Sender: news@cs.vu.nl
Organization: Fac. Wiskunde & Informatica, VU, Amsterdam
Date: Mon, 5 Feb 1996 13:49:14 GMT
X-Newsreader: knews 0.9.4
Message-ID: <DMB2E3.A13.0.-s@cs.vu.nl>
Lines: 33

In article <4f1997$8c6@ctylnk.cityu.edu.hk>,
	90853590@cpccux0.cityu.edu.hk () writes:
>Hello, everybody,

// CUT

>
>1.	Why BP is the most usual choice?

Mainly because of historical reasons. BP WAS the best and most used algorithm.
Meanwhile there are a lot of other algorithms, which suit better and are
faster for specific tasks.

>2.	Does the architecture determine the suitability of the ANN for different applications?

Conclusion: Yes, BUT I didn't encounter any specific or suitable architecture
for stack market forecasting or analysis. If you know or get to know anything
in that direction I really am very interested in that field of ANN-ing.

>
>Regards,
>KK LEUNG
>


A.G. ten Oever







