Newsgroups: comp.ai.neural-nets,comp.ai.genetic,comp.ai.alife
Path: cantaloupe.srv.cs.cmu.edu!bb3.andrew.cmu.edu!newsfeed.pitt.edu!godot.cc.duq.edu!news.duke.edu!news.mathworks.com!newsfeed.internetmci.com!in1.uu.net!bcstec!nntp
From: Mike Fouche <mikef@hsvaic.hv.boeing.com>
Subject: Re: GA for learning in ANN
X-Nntp-Posting-Host: e570254.hv.boeing.com
Content-Type: text/plain; charset=us-ascii
Message-ID: <DqoyKy.L52@bcstec.ca.boeing.com>
To: om@alto.unice.fr
Sender: nntp@bcstec.ca.boeing.com (NNTP News Access)
Content-Transfer-Encoding: 7bit
Organization: Missiles & Space Division
References: <4m2njb$btc@malibu.unice.fr>
Mime-Version: 1.0
Date: Tue, 30 Apr 1996 20:07:45 GMT
X-Mailer: Mozilla 1.2 (Windows; U; 16bit)
Lines: 12
Xref: glinda.oz.cs.cmu.edu comp.ai.neural-nets:31349 comp.ai.genetic:8629 comp.ai.alife:5515

Hello,

From what I've seen of GAs, they might be suited for very large networks 
where a traditional optimization method would bog down.  However for 
typical applications I'd think that standard methods are faster.  We use
a Boeing-proprietary method that usually converges from several seconds 
to several minutes to an accumulated output error of 1e-15.  I seriously 
doubt that GAs can match that performance.

Mike Fouche
Boeing Missiles & Space Division

