
Genetic Algorithms Digest   Thursday, July 16 1992   Volume 6 : Issue 25

 - Send submissions to GA-List@AIC.NRL.NAVY.MIL
 - Send administrative requests to GA-List-Request@AIC.NRL.NAVY.MIL
 - anonymous ftp archive: FTP.AIC.NRL.NAVY.MIL (see v6n5 for details)

Today's Topics:
	- Re: a specific crossover operator closure problem (v6n24)
	- Mirror Archive for GA-list
	- release announcement: GAucsd 1.4
	- Need info on GA in Human Factors
	- GA Object Oriented Database Environment
	- Looking for advice on Michigan & Pitts. Approaches

****************************************************************************

CALENDAR OF GA-RELATED ACTIVITIES: (with GA-List issue reference)

 10th National Conference on AI, San Jose,                      Jul 12-17, 92
 FOGA-92, Foundations of Genetic Algorithms, Colorado (v5n32)   Jul 26-29, 92
 COG SCI 92, Cognitive Science Conf., Indiana, (v5n39, v6n22)   Jul 29-1,  92
 ECAI 92, 10th European Conference on AI (v5n13)                Aug  3-7,  92
 Parallel Problem Solving from Nature, Brussels, (v5n29)        Sep 28-30, 92
 SAB92, From Animals to Animats, Honolulu (v6n6)                Dec  7-11, 92
 ICNN93, IEEE Intl. Conf. on Neural Networks, Calif (v6n24)     Mar 28-1,  93
 Intl. Conf. on Neural Networks and GAs, Innsbruck (v6n22)      Apr 13-16, 93

 (Send announcements of other activities to GA-List@aic.nrl.navy.mil)

****************************************************************************
----------------------------------------------------------------------

From: vanlent@AIC.NRL.Navy.Mil
Date: Tue, 14 Jul 92 09:46:18 EDT
Subject: Re: a specific crossover operator closure problem (v6n24)

   In response to "a specific crossover operator closure problem," (v6n24):

   I think that this is the very same problem that comes up when working on
   traveling salesman problems and representing each tour as the list of
   cities in the order they are visited.  You have a long integer tuple in
   which no integer(city) can be repeated.

   There have been a fair number of suggested crossovers for this type of
   genome.  Some popular ones are Davis's Modified Crossover, Goldberg and
   Lingle's PMX, and Oliver, Smith, and Holland's Cycle crossover all of
   which are described in the Oliver, Smith, and Holland paper in the
   Proceedings of the 2nd. Int.  Conference on GA's.  There is also the edge
   recombination operator described in Whitley's paper in the 3rd. conference
   proceedings and Muhlenbein presents another approach in the proceedings of
   the Foundations of Genetic Algorithms workshop from 1990.  All of these
   approaches suffer from the fact that they are not true crossover in that
   they introduce some variation as they recombine the genomes.

   Another approach is to find some different way to encode your genome such
   that a normal crossover is legal.  Grefenstette shows how to do this for
   the TSP in his paper in the proceedings of the 1st conference on Genetic
   Algorithms.  While his exact approach may not work exactly for your
   problem something similar might.

   Hope all this is a little helpful.

   Mike van Lent (vanlent@aic.nrl.navy.mil)
   AI Center at the Naval Research Lab

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

From: sourav@archsci.arch.su.OZ.AU
Date: Sun, 12 Jul 1992 20:29:46 +1000
Subject: Mirror Archive for GA-software.

   We have recently formed a Discussion Forum for Genetic Algorithms to bring
   together the GA-researchers in Australia. It was announced in "GA-List
   v6n23".  The forum is called "AUSGA". We have 24 members now !

   We have a mirror site for archieving all the GA-software available at
   "ftp.aic.nrl.navy.mil".  The site is "archsci.arch.su.oz.au". The
   directory is "/pub/ausga". This directory has already been opened and I
   will be in charge of its maintenance and updating.

   This will essentially save us a lot of satellite bandwidth as these
   GA-softwares are getting more and more popular in Australia and around the
   world !

   The researchers and students working with GA should now be able to fetch
   all the source-codes of various GA-softwares via anonymous ftp to
   "archsci.arch.su.oz.au". This site will mirror the contents of
   "ftp.aic.nrl.navy.mil".

   -sourav-

   ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   | SOURAV KUNDU                                         |
   | Research Staff; PhD student                          |
   | Design Computing Unit                                |
   | Department of Architectural and Design Science (G04) |
   | University of Sydney                                 |
   | Sydney, NSW 2006. Australia                          |
   |                                                      |
   | E-Mail    sourav@archsci.arch.su.oz.au               |
   | Tel:     +61-2-692 2053                              |
   | Fax:     +61-2-692 3031                              |
   ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

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

From: schraudo@cs.UCSD.EDU (Nici Schraudolph)
Date: Fri, 10 Jul 92 19:03:54 PDT
Subject: release announcement: GAucsd 1.4

   Version 1.4 (yes, we skipped 1.3) of the GAucsd GA software has now been
   released.  The basic functionality is identical to the previous version,
   but a number of bugs have been fixed, the documentation has had a work-
   over, and the portability of the code has been improved.  GAucsd 1.4 is
   now easily ported to most systems that have "make", "awk" and a reasonable
   C compiler; limited environments such as DOS are the primary beneficiaries.

   As before, GAucsd 1.4 and related material is available via anonymous ftp
   from cs.ucsd.edu (132.239.51.3) in the pub/GAucsd directory.  The file
   GAucsd14.sh.Z contains the source distribution in compressed shar format,
   including an ASCII version of the User's Guide.  The compressed PostScript
   version is available in the same directory as GAucsd14.ps.Z.  The source
   distribution is also available from my e-mail server: send me an EMPTY
   message with the SUBJECT line containing the string "send GAucsd source".

   Note that the related article on Dynamic Parameter Encoding has appeared
   in the current issue of Machine Learning (9:9-21) and has therefore been
   removed from the ftp server.  Reprints of this article (technical report
   CS90-175) as well as the GAucsd User's Guide (technical report CS92-249)
   can be obtained for a charge of 5 US$ (prepaid) from the CSE Department,
   U.C. San Diego, La Jolla, CA 92093-0114, USA.

   - Nici Schraudolph.


   [Editor's Note: GAucsd 1.4 is also available from ftp.aic.nrl.navy.mil.
   The file is: /pub/galist/source-code/ga-source/GAucsd14.tar. -- ACS]

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

From: toet@izf.tno.nl (Lex Toet)
Date: Thu, 9 Jul 92 15:24:51 +0200
Subject: Need info on GA in Human Factors

   Does anyone know of applications of Genetic Algorithms to Human Factors,
   where the objective function is replaced by the human expert.

   Any information will be greatly appreciated.

   Lex Toet
   Institute for Perception TNO
   The Netherlands
   toet@izf.tno.nl

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

From: objy!server!chris@Sun.COM
Date: Fri, 10 Jul 92 14:35:24 -0700
Subject: GA Object Oriented Database Environment

   I am interested in building some GA tools on top of Objectivity/DB.

   This would provide a C++ application langauge, and large-scale object
   database facilities.

   Questions:

   Does anyone have experience to contribute to this endeavor?  What are some
   features that might be useful when considering a database approach?  What
   type of projects now make more sense with this persistence capability? eg.
   A history trail of evolutionary stages for successful chromosomes?

   Others?

   thanks, Chris

   Objectivity Inc.                                            Chris Rigatuso 
   800 El Camino Real, 3th floor                               chris@objy.com
   Menlo Park, CA 94025                                  ...!uunet!objy!chris
   USA                      fax: +1 415 325 0939              +1 415 688 8070

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

From: Gihan Seneviratne <Gihan.P.Seneviratne@computing-maths.cardiff.ac.uk>
Date: Tue, 14 Jul 92 10:28:26 BST
Subject: Looking for advice on Michigan & Pitts. Approaches

   Hi,

	   I am looking for detailed references (or/and advice) on 
	   "Michigan Approach" and "Pitts. Approach" in Classifier 
	   systems.
	   I would appreciate any advice or pointers to advice 
	   regarding this.
	   Please respond via email.

	   Thanks in advance.

    .............................................
   . GIHAN P. SENEVIRATNE                        .
   . Research Student                            .
   . Department of Computing Mathematics,     ................................
   . University of Wales College of Cardiff, /| e-mail: <scmgps@cm.cf.ac.uk> |\
   . Mathematics Building,                  { | Fax   : 0222-666182          | }
   . Senghennydd Road, CARDIFF               \|..............................|/
   . CF2  4AG                                    .
   .                                             .
   . UNITED KINGDOM                              .
   <.............................................>


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