Genetic Algorithms Digest    Thursday, 28 September 1989    Volume 3 : Issue 14

 - Send submissions to GA-List@AIC.NRL.NAVY.MIL
 - Send administrative requests to GA-List-Request@AIC.NRL.NAVY.MIL

Today's Topics:
	- MLJ special issue on GAs
	- Re: GA for Scheduling
	- Re: Partial matching in classifier systems

--------------------------------

Date: Thu, 28 Sep 89 10:26:11 EDT
From: Kenneth Dejong <dejong@aic.nrl.navy.mil>
Subject: MLJ special issue on GAs

As announced at ICGA-89, I will be the editor of the second special issue
on Genetic Algorithms for the Machine Learning Journal (Kluwer).  The page
constraint of approximately 100 pages/issue suggests a default model of
four papers of about 25 pages each.  In addition, the content of the papers
should be of interest to the machine learning community.  If there are more
than 4 quality papers that fit these constraints, I have been assured that
they can be accomodated (via a double issue, a later third issue, in a
normal issue, etc.).  My goal is to to have the special GA issue appear
as the third MLJ issue of 1990.  In order to do that, we need to meet the
following schedule:

	October 20:	rough draft submitted to me for early feedback on
			suitability, etc.

	November 20:	reviewable versions submitted to me for final
			evaluation and feedback.

	December 20:	papers returned with reviewers comments and
			final decision.

	January 20:	final versions to me.

	February 20:	complete issue to Kluwer.

If you are an interested author, send me a note to that effect and
start working on that first draft!

Ken De Jong	dejong@aic.nrl.navy.mil
		kdejong@gmuvax2.gmu.edu


--------------------------------

Date:     Tue, 26 Sep 89 11:02:31 GMT
From: pat%computer-science.strathclyde.ac.uk@NSFnet-Relay.AC.UK
Subject:  Re: GA for Scheduling

Mark Hughes (v3n13) writes:

> I will be grateful for any anecdotes/references for work on factory
> scheduling using GAs.  So far I'm aware of two papers; one by Davis
> (1985) and one by Hilliard et al (1987), from the first and second GA
> conference procs. respectively. (These are on order).

See also:

Derek Smith, "Bin packing with Adaptive Search"
Proceedings of ICGA 1985

Patrick Prosser, "A Hybrid Genetic Algorithm for Pallet Loading",
European Conference on Artificial Intelligence, Munich 1988

--------------------------------

Date: Wed, 27 Sep 89 07:50:52 EDT
From: booker@AIC.NRL.NAVY.MIL
Subject: Re: Partial matching in classifier systems

> Dave Goldberg gives in his book "Genetic Algorithms in Search, Optimization
> and Machine Learning" a brief discussion of partial matching is given (pp.
> 277-278). One of the things he states there is that the original partial
> match score used by Booker in his thesis [Booker 1982]
> has been critizied and due to that criticism Booker has introduced other
> match scores in his ICGA85-paper [Booker 85].
> Does anybody know what these critics are?

The differences in the partial match score in my thesis
and the one in my ICGA85 paper have nothing to do with
criticisms of partial matching. The latter version simply
works better.

> Goldberg states also that a lot of people oppose partial matching.
> 
> Can anybody tell me what the arguments of the opposants are against partial

The only objections to partial matching I'm aware of
argue that it is an unnecessary complication to the
basic execution cycle. The claim is that "create"
operators are adequate for managing situations where
no rule matches exactly. I obviously disagree, and I
think I've shown that partial matching can be used
in a limited way that doesn't complicate the basic cycle.

Lashon

--------------------------------

End of Genetic Algorithms Digest
********************************
