Genetic Algorithms Digest    Tuesday, 5 July 1988    Volume 2 : Issue 16

 - Send submissions to GA-List@NRL-AIC.ARPA
 - Send administrative requests to
	John Grefenstette <GA-List-Request@NRL-AIC.ARPA>

Today's Topics:
	- Goldberg Book
	- Society of Mind
	- GAs vs Evolutionary Programming
	- GAs and Neural Networks
	- Next GA Conference

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

Date:         Fri, 10 Jun 88 16:22:42 CST
From: Dave Goldberg <DGOLDBER@ua1vm.bitnet>
Subject:      GASOML ETA

My book entitled GENETIC ALGORITHMS IN SEARCH, OPTIMIZATION, AND MACHINE
LEARNING will be bound and available on August 1, 1988 (really).  It will
be available from Addison-Wesley and may be ordered now by mail.  The book
is an introductory text and self-study guide.  It contains worked out examples,
analysis, and even some programs.  Any GA-List subscriber interested in
advance ordering information should send me an e-mail note with an r-mail
(real mail) address so I can send a descriptive flyer. Dave Goldberg

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

Date: Sat, 11 Jun 88 14:41:59 EDT
From: bill coderre and his pets <bc@media-lab.media.mit.edu>
Subject: Society of Mind

I am currently working on getting Societies of Mind to learn.

It seems an extension to the idea behind Genetic Algorithms holds
promise.

In order to perform selection, let's have win/lose detectors attached
to agents. When a detector fires, we need to do some credit/blame
assignment to some sub-agents. This knowledge is situational: "Agent
78 had a good idea about how to eat corn on the cob".

In order to perform mutation, we need an agent generator that can
understand the situational knowledge stored in the agents and create
new sub-agents heuristically.

Any comments??

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

Date: Fri, 10 Jun 88 08:40 PDT
From: nesliwa%nasamail@ames.arc.nasa.gov (NANCY E. SLIWA)
Subject: GA vs Evolutionary Programming

Has anyone compared genetic algorithms with Dr. Larry Fogel's concept of
"evolutionary Programming"? It would be interesting to examine the differences.

Nancy Sliwa
NASA Langley

[ The main differences between GAs and Fogel's work is that
Fogel used a population size of 1 and relied on mutation as the
primary search operator, resulting in a technique closer to
simulated annealing than to GAs. -- JJG ]

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

Date: Fri 10 Jun 88 12:39:37-EDT
From: DDAVIS@g.bbn.com
Subject: Genetic Algorithms and Neural Networks

I have noticed several queries on the relation between neural networks
and genetic algorithms posted to the ga-list lately.  I am working on a
paper that describes a hybrid system that accommodates both sorts of
learning automata.  My paper also describes the "projection" of each
sort of automaton onto the plane of the other.  When I complete the
paper (in a month or two) I will post an abstract and announcment of
where it will be given.

Regards,

David Davis.

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

Date: Tue, 14 Jun 88 19:15:08 pdt
From: David Rogers <drogers@icarus.riacs.edu>
Subject: Request for info about next GA conference

Do you happen to know where/when the next GA conference will be?

Thanks in advance,

David Rogers
drogers@riacs.edu


[ Tentative plans are to have the Third International Conference
on Genetic Algorithms and Their Applications in Washington, DC,
in June or July of 1989.  Exact location and dates will be announced
shortly.  Ken De Jong will serve as Conference Chair and Dave Schaffer
will be Program Chair.  More details will appear here as they become
available. -- JJG]

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

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