Newsgroups: comp.ai.neural-nets,comp.ai
Path: cantaloupe.srv.cs.cmu.edu!rochester!udel!news.mathworks.com!uhog.mit.edu!bloom-beacon.mit.edu!news.starnet.net!wupost!howland.reston.ans.net!usc!sdd.hp.com!hp-pcd!hpcvsnz!briand
From: briand@cv.hp.com (Brian Dixon)
Subject: Re: Neural nets: the argument from evolution
Sender: news@hpcvsnz.cv.hp.com (News )
Message-ID: <D9t513.6Jr@hpcvsnz.cv.hp.com>
Date: Wed, 7 Jun 1995 14:46:14 GMT
References: <3q2oi9$11n@nic.wi.leidenuniv.nl>
Nntp-Posting-Host: hpcvsgen.cv.hp.com
Organization: Hewlett-Packard
X-Newsreader: TIN [version 1.1 PL9.4]
Followup-To: comp.ai.neural-nets,comp.ai
Lines: 15
Xref: glinda.oz.cs.cmu.edu comp.ai.neural-nets:24570 comp.ai:30446

E.J.W. Boers (boers@wi.leidenuniv.nl) wrote:

: I think that in order to evolve really interesting behaviour of
: artificial agents, it is necessary to have several agents interacting
: with eachother. An element of competition between the different agents,
: combined with the need to cooperate to survive, will in my opinion
: be the best approach. 

You mean like competition leading to a) natural selection, genetic interaction
like b) crossover, and c) mutation...then go back to a), etc.?  Hmmmmm, sounds
like GA to me...

--
Brian Dixon, Machine Vision Engineer, Hewlett Packard (Corvallis, Oregon)
503-715-3143 (wk), briand@cv.hp.com (email). "Opinions & attitudes are mine!"
