Newsgroups: comp.ai
Path: cantaloupe.srv.cs.cmu.edu!bb3.andrew.cmu.edu!nntp.sei.cmu.edu!news.cis.ohio-state.edu!magnus.acs.ohio-state.edu!csn!nntp-xfer-1.csn.net!su-news-hub1.bbnplanet.com!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!cam-news-hub1.bbnplanet.com!howland.erols.net!feed1.news.erols.com!news.ecn.uoknor.edu!solace!mn6.swip.net!plug.news.pipex.net!pipex!cold.news.pipex.net!pipex!dish.news.pipex.net!pipex!NEWS!not-for-mail
From: Rob Holmes <rob.holmes@powertech.co.uk>
Subject: Re: Neural Networks.
Message-ID: <32D2528D.57D0@powertech.co.uk>
Date: Tue, 07 Jan 1997 13:41:33 +0000
References: <32C25D2A.2FA7@aracnet.net>
X-Mailer: Mozilla 3.0 (Win95; I)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Lines: 18

eduardo montana wrote:
> 
> I am currently searching for:
> Neural software that supports the following:
> Fast, industrial strength, mission-critical neural network development
> and implementation to run on Windows NT (preferably with genetic
> optimisation).

Ed, I've been using Neuralware's product Neuralworks for a couple of
years now and it seems pretty well behaved.  I've been using it in the
way you describe: embedding the produced C into a hybrid systems which
gets delivered to the customer.  I believe the current version (v 2.51?)
does have genetic optimisation in but its not a feature I've used.

One thought - my experience with forecasting is that it is particularly
important to get the preprocessing right and ensure that I've extracted
as much of the info using standard linear techniques (the Holt-Winters
procedure, for example) before resorting to the non-linear.

