Re: GP and Swarm

[goellner]

Yep, once again I crossed up my terms. I have seen tools for GA activities
which is what my original post should have said. I was looking for a
response concerning tools for GP.

As you pointed out GA and GP are different and it is easy to confuse the
terms. But they are quite different approaches.

I mentioned LISP because I know LISP and I don't really care to learn any
more TCL than I have to. I disagree with your comment that GA's can do
most of the stuff that GP's can do. The most that can be said is that they
have intersecting problem domains. 

-------
Chris Goellner
UNIX Systems Admin
Ciba Specialty Chemicals

On Wed, 3 Sep 1997, J.J. Merelo Guervos wrote:

> >>>>> "bleydorf" == bleydorf  <address@hidden> writes:
> 
> Hi,
> 
>     bleydorf> Isn't the Breeder stuff(by JJ Merelo) a genetic library?
> 
> Yep, but GP and GA are somewhat different. GP is a way of evolving
> LISP programs, that doesn't use mutation in the same way that GA does,
> and uses specialized operators to crossover LISP S-expressions. 
> 
>     bleydorf> I have downloaded it and planned to explore it in the
>     bleydorf> near future.  I checked out Lawrence Davis's book
>     bleydorf> Guidebook on Genetic Programming and it looks like
> 
> Is'nt that Handbook of Genetic Algorithms?...
> 
> @Book{Davis91,
>   author =       "Lawrence Davis",
>   title =        "Handbook of Genetic Algorithms",
>   publisher =    "Van Nostrand Reinhold",
>   year =         "1991",
>   address =      "New York",
> }
> 
> For references on genetic programming, check out
> http://www.geneticprogramming.com. 
> 
>     bleydorf> Breeder definately contains genetic "stuff"...
> 
> Yep, but it cannot be used to evolve LISP, or any other kind of
> programs, since it's chromosomes are binary strings, and not LISP
> expressions.
> 
> As for using genetic programming in Swarm, it would probably be a
> matter of embedding a LISP interpreter into it. I don't know if it's
> really worth the while, most of the stuff that can be done with GP can
> be done with GA too. There is already an interpreter embedded in
> Swarm, tcl interpreter, but I doubt somebody has tried to make tcl
> programs evolve. What makes LISP S-expressions evolvable is (found in
> page 71 of Koza's "Genetic programming") probably makes tcl evolvable
> too. Most of the seven reasons Koza offers for LISP are check out for
> tcl too. Well, just a thought.
> 
>                               JJ
> 
> PS: Just did a GA in JavaScript, in case somebody is interested: check
> it out at http://kal-el.ugr.es/~jmerelo/GAJS.html.
> 
> -- 
> JJ Merelo                         | http://kal-el.ugr.es/htbin/jj-plex
> Grupo Geneura ---- Univ. Granada  | http://kal-el.ugr.es/
> 
> 
>                   ==================================
>    Swarm-Support is for discussion of the technical details of the day
>    to day usage of Swarm.  For list administration needs (esp.
>    [un]subscribing), please send a message to <address@hidden>
>    with "help" in the body of the message.
>                   ==================================
> 


                  ==================================
   Swarm-Support is for discussion of the technical details of the day
   to day usage of Swarm.  For list administration needs (esp.
   [un]subscribing), please send a message to <address@hidden>
   with "help" in the body of the message.
                  ==================================