ANN: pygene - genetic algorithms package
Erik Max Francis
max at alcyone.com
Tue Dec 6 21:48:11 EST 2005
Peter Hansen wrote:
> I've done just enough work in genetic algorithms (and a token amount in
> genetic programming) to be perplexed by this comment. Are you
> suggesting that genetic programming is somehow not related to genetic
> algorithms?
>
> My understanding is that (said perhaps somewhat simplistically) genetic
> programming is an application of genetic algorithms in which the genomes
> are treated as describing the structure of a program whose execution
> basically results in the fitness level for that genome.
>
> If that's a reasonably accurate statement (or, I suppose, even if it's
> not), would you please clarify how your "fixed genome" comment relates
> to any of this?
You're not replying to me, but I'm the one that elicited that comment.
(I was originally asking the question because I misinterpreted the first
sentence of his announcement about pygene to mean that pygene was a
genetic programming system, but that was never his claim.)
A genetic algorithm involves manipulating "programs" which consist of a
fixed amount of homogeneous data, for instance, an array of neural
network weights, or the coefficients to an equation. Genetic
programming involves manipulating general programs, usually as some form
of tree. The classic model for genetic programming, from Koza, is where
the programs to be manipulated are Lisp s-expressions.
pygene implemented a genetic algorithm system, not genetic a programming
system, hence his response. It was only my interpretation of his
introductory comment that led anyone to believe otherwise.
--
Erik Max Francis && max at alcyone.com && http://www.alcyone.com/max/
San Jose, CA, USA && 37 20 N 121 53 W && AIM erikmaxfrancis
Men live by forgetting -- women live on memories.
-- T.S. Eliot
More information about the Python-list
mailing list