ftp.cs.arizona.edu/ftp.cs.arizona.edu.tar/ftp.cs.arizona.edu/icon/newsgrp/group96b.txt/000068_icon-group-sender _Wed Dec 11 13:17:02 1996.msg

Internet Message Format | 1997-01-02 | 2KB

Received: by cheltenham.cs.arizona.edu; Wed, 11 Dec 1996 12:32:06 MST
Sender: [e-mail]
Message-Id: <[e-mail]>
Date: Wed, 11 Dec 1996 13:17:02 -0600
From: Ray Pereda <[e-mail]>
Organization: Texas Instruments
X-Mailer: Mozilla 3.0Gold (X11; I; SunOS 4.1.4 sun4m)
Mime-Version: 1.0
To: [e-mail]
Cc: [e-mail]
Subject: Re: Genetic Algorithms
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Errors-To: [e-mail]

Hello Roger,

I've used Icon to write three large genetic algorithms programs.  Much
of my code is based on the design and code from David Goldberg's book on
genetic algorithms.  Icon's strings are very easy to use for the genetic
encoding of one's problems.  The C and Pascal implementations in
Goldberg's book are about three times as long as the Icon implementation
that I developed.  I can give you a copy of the code.  If other's are
interested, I can make it more generally available.

I used Icon to research basic properties of GAs.  I can email you a copy
of the paper detailing the research.  I've also written super hand
optimized C programs to do this research.  What I found was that the C
program took so much more effort to tinker with different ideas that I
eventually abandoned it.  Instead, I worked with Icon programs that are
twice as slow but 20 times easier to modify.  When you're doing lots of
tinkering, 20 times easier makes a huge difference.

-Ray Pereda
[e-mail]

> Date: 11 Dec 96  09:50:19 GMT
> From: R J Hare <[e-mail]>
> Subject: Genetic algorithms
> To: [e-mail]
> Reply-To: [e-mail]
> Organisation:  Edinburgh University Computing Services
>
> Has anyone got any experience of using Icon to program genetic
> algorithms? Are they prepared to share that experience?
>
> Thanks
>
> Roger Hare