Newsgroups: comp.ai.neural-nets
Path: cantaloupe.srv.cs.cmu.edu!rochester!cornellcs!newsstand.cit.cornell.edu!news.tc.cornell.edu!news3.cac.psu.edu!howland.erols.net!www.nntp.primenet.com!nntp.primenet.com!netcom.com!nagle
From: nagle@netcom.com (John Nagle)
Subject: Re: GAs: breakthrough or just another optimization strategy ?
Message-ID: <nagleE1zBFr.Lw8@netcom.com>
Organization: NETCOM On-line Communication Services (408 261-4700 guest)
References: <32A2BF61.41C67EA6@robot.uvsq.fr> <586adl$mt9$1@mhafc.production.compuserve.com>
Date: Fri, 6 Dec 1996 06:31:03 GMT
Lines: 16
Sender: nagle@netcom6.netcom.com

Will Dwinnell <76743.1740@CompuServe.COM> writes:
>It is worth exploring further which situations are generally more 
>amenable to solution by one method vs. another.  One great 
>advantage of the newer optimization systems (GAs, SA, etc.) is 
>that they are "blind" in the sense that they do not require 
>knowledge of the problem domain nor do they require computation 
>of the utility function's gradient.

     Actually, GAs do have certain implicit assumptions about the
nature of the space being searched, although it's hard to describe
what they are.

     On the downside, where there is continuity, GAs don't exploit it
very well.

					John Nagle
