home *** CD-ROM | disk | FTP | other *** search
- Comments: Gated by NETNEWS@AUVM.AMERICAN.EDU
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!howland.reston.ans.net!paladin.american.edu!auvm!UIUC.EDU!G-CZIKO
- Return-Path: <@VMD.CSO.UIUC.EDU,@vmd.cso.uiuc.edu:g-cziko@uiuc.edu>
- X-Sender: cziko@ux1.cso.uiuc.edu
- Message-ID: <199301121655.AA06672@ux1.cso.uiuc.edu>
- Newsgroups: bit.listserv.csg-l
- Date: Tue, 12 Jan 1993 10:55:45 -0600
- Sender: "Control Systems Group Network (CSGnet)" <CSG-L@UIUCVMD.BITNET>
- From: "CZIKO Gary A." <g-cziko@UIUC.EDU>
- Subject: Koza, Genetic Programming, & LISP
- Lines: 57
-
- [From Gary Cziko 930112.1620 GMT]
-
- Greg Williams (930112) says:
-
- >Buried under two feet of white paper, rather than two feet of white H2O. Gary,
- >Koza's mammoth tome didn't help this situation. Pat's first comment was,
- >"Yeah, LISP -- it makes sense for this. Or assembly language (like in
- >TIERRA)." Mine: "Somebody's going to make a lot of bucks off this for a long
- >time." We're now thinking that some of the ideas will be helpful in writing
- >our "super" NSCK (with reorganization). I recommend, especially to Martin
- >Taylor and his colleagues, John R. Koza, GENETIC PROGRAMMING: ON THE
- >PROGRAMMING OF COMPUTERS BY MEANS OF NATURAL SELECTION, MIT Press, >1992.
-
- Greg, now you know I've spent just about every "free" minute over the last
- week with Koza's book. And I am not even a computer programmer! I never
- dreamed using "natural" selection to evolve computer programs could be so
- simple and understandable (even for me) and that it would so quickly result
- in highly fit programs (usually within about 30 generations of a population
- of 500 programs) for quite complex problems. I just wish Koza had said
- more about the weaknesses and failures of this appproach. He makes it look
- like there aren't any. Maybe that's the case, but I wonder.
-
- Can you say a few words about the LISP vs. assembly language (Tierra)
- approach to evolving computer programs.
-
- Also, would it be hard to simulated a problem with disturbances and see
- control systems and hierarchies thereof evolve? One could start with the
- basic tracking task.
-
- One thing that has be puzzled, however, is the "money" aspect of all this.
- Koza includes lots of LISP source code in his book which make it look like
- he wants others to use his techniques. But then I see in the references
- that he has patented this techniques. What does that mean? We can use the
- code for academic purposes but not for making money? I can understand how
- software can be copyrighted, but I've never run into patents before
- concerning computer programs.
-
- Greg, can you or anybody else in CSGnet land tell me some more about LISP?
- I'd love to play around with the "genetic" techniques Koza uses. But if I
- take the trouble to learn LISP, will I also be able to do with the stuff
- that you, Bill, and Rick do with C and Pascal? Are there ways of
- converting LISP code to C or Pascal to make LISP programs more portable?
-
- >(And
- >also, by the way, Robert A.M. Gregson, TIMES SERIES IN PSYCHOLOGY, Erlbaum,
- >1983.)
-
- Can you tell us why you recommend this?--Gary
- ------------------------------------------------------------------
- Gary A. Cziko Telephone: 217-333-8527
- Educational Psychology FAX: 217-244-7620
- University of Illinois E-mail: g-cziko@uiuc.edu
- 1310 S. Sixth Street Radio: N9MJZ
- 210 Education Building
- Champaign, Illinois 61820-6990
- USA
- ------------------------------------------------------------------
-