home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
ftp.cs.arizona.edu
/
ftp.cs.arizona.edu.tar
/
ftp.cs.arizona.edu
/
icon
/
newsgrp
/
group93b.txt
/
000082_icon-group-sender _Thu May 6 10:22:43 1993.msg
< prev
next >
Wrap
Internet Message Format
|
1993-06-16
|
2KB
Received: by cheltenham.cs.arizona.edu; Tue, 11 May 1993 21:03:24 MST
Date: 6 May 93 10:22:43 GMT
From: cis.ohio-state.edu!zaphod.mps.ohio-state.edu!swrinde!sdd.hp.com!sgiblab!munnari.oz.au!cs.mu.OZ.AU!mundoe!foda@ucbvax.Berkeley.EDU (Omar Foda)
Organization: Computer Science, University of Melbourne, Australia
Subject: icon in genetic programming
Message-Id: <foda.736683763@mundoe>
Sender: icon-group-request@cs.arizona.edu
To: icon-group@cs.arizona.edu
Status: R
Errors-To: icon-group-errors@cs.arizona.edu
Hello, I am vaguely familiar with icon: I still do not
have the book, but I would still like to pose the following
question:
I am interested in genetic programming a la Koza. Koza's
pioneering code was written in lisp, basically because
a lisp program is its own parse tree: in other words, the
parse tree is available to the programmer to 'genetically
manipulate'. I would like to do the same thing, at the
level of Mathematica functions which can also be written
consistently in terms of binary trees: I wish to produce the best
possible solution to a Mathematical problem in terms of
a function that can be evaluated by Mathematica. The starting
point would be a population of randomly generated, but
syntactically correct composite Mathematica functions.
These will have to be cross-bred, mutated, etc. in such
a way that the resulting functions are also correct functions.
This will require extensive pattern scanning, substitution,
exchanging, etc. something that can be done in Mathematica,
but extremely slowly. I can probably do that also in awk.
My question is: is that something that can be easily done
in icon? More interestingly for me: are there others who
are busy with similar problems?
Thank you in advance, Omar Foda.