Comments for Pragmatic Genetic Programming

Comment on Eureqa II, aka Formulize by bill

bill — Fri, 19 Oct 2012 12:12:04 +0000

Samira, you should probably look at ECJ for an open-source GP project with a focus on symbolic regression.

Comment on Eureqa II, aka Formulize by samira rezai

samira rezai — Wed, 17 Oct 2012 13:30:09 +0000

Hi
I am a M.S student and I choose symbolic regression via GP for my project, but unfortunately I can’t find good source especially about the SR and it’s importance and features and … about it. I really need help. please help me and introduce good sources.
thanks

Comment on Pizza Eating Strategies, part 1 by Pizza Eating Strategies, Part 2 | Pragmatic Genetic Programming

Pizza Eating Strategies, Part 2 | Pragmatic Genetic Programming — Thu, 23 Aug 2012 11:04:32 +0000

[…] Post navigation ← Previous […]

Comment on The Three Languages of GP by Pizza Eating Strategies, part 1 | Pragmatic Genetic Programming

Pizza Eating Strategies, part 1 | Pragmatic Genetic Programming — Sat, 18 Aug 2012 13:16:50 +0000

[…] to use Lee Spector’s Tag Space Machine system as an architecture in my Answer Language ; […]

Comment on Applying GP to Winkler’s Pizza Eating Game by Pizza Eating Strategies, part 1 | Pragmatic Genetic Programming

Pizza Eating Strategies, part 1 | Pragmatic Genetic Programming — Sat, 18 Aug 2012 13:10:22 +0000

[…] Post navigation ← Previous […]

Comment on Eureqa II, aka Formulize by Matías

Matías — Tue, 17 Jul 2012 01:57:16 +0000

Thank you very much for the tip,
wasn’t acquainted with the program
and now that I downloaded it, I find it pretty useful
for simple exploratory analysis and little prediction problems
even if I have now idea how to model them (but when you do, it also can use that knowledge quite nicely).

Comment on On sensibility by bill

bill — Sun, 08 Jul 2012 22:18:10 +0000

Jason: It's an interesting thought, but I think the difficulty is deeper by far than folks needing to be reminded of context by proximity. The core of the culture is really something like: you must have an idea of your own, and you must work on it on your own until it has made a "significant contribution" to the community—which might be translated as "not obvious for the skilled person in view of the state of the art in general". To be honest, I suspect any specific and lively field would start to claim GP as "its own" as soon as they both started paying productive attention to one another. And that might be OK. JJ: All of us solve problems because they're there. That's something I absolutely agree with, and be careful not to misunderstand my call to construct a narrative that provides context for one's work with the demand that you need that context before the work is done. What I'm saying is that these researchers should be allowed to say they did it because it was what was needed to get to their goal, or even that it was sufficient for their goal. We use crossover because we had to use something. What is unproductive, and what I (and an awful lot of reviewers mention to me as we chat) am complaining about is the stories we are told (in papers and talks) that after the fact read like, "I made this thing. Look, it's better than the old one." Better for what? For whom? Why was it made? Inspiration is hidden, and in turn usefulness is hidden, when the "story" is hollow in this way. The subject can be theory, can be subjective, can be "real" or "practical"—but there needs to be some subject when we hear the "result". And yes, I'm well aware that work can be used exaptively and in innovative contexts. And I'm aware of Évariste Galois and his infamous duel, and how his work is now so important. But if he had won his duel, he would have been able to tell that story; because he told no story, it languished, obscure.

Comment on On sensibility by JJ

JJ — Sun, 08 Jul 2012 15:38:48 +0000

The first answer to why you are solving a problem is because it's there. The second, to why you did it that particular way, is because it seemed a good idea. You'll have to look at half a dozend Q&A to arrive at something resembling "The customer requested it". I don't think it's a good idea either to do "customer-driven" science, but, at the end of the day, solving an academic problem you acquire an expertise that, eventually, will help you mix the perfect cocktail or create the next Facebook. After all, Neal Stephenson used evolutionary algorithms as a trope for distributing the, and that's my English failing me here, heirloom of an old aunt among a big family with many different requirements. So they are not excuses, it's just the old SEB, somebody else's business: technology transfer departments in the universities, for instance, or marketing in companies.

Comment on On sensibility by Jason H. Moore, Ph.D.

Jason H. Moore, Ph.D. — Sun, 08 Jul 2012 13:49:32 +0000

Bill,

Thanks for sharing your thoughts. I don’t agree with everything you said but do think you made some good points. The bottom line is that this is a computer science conference. To address some of your concerns I would recommend GECCO be held in conjunction every year with other domain-specific conferences. For example, ISMB (http://www.iscb.org/ismb2012/) is being held next week. This is one of the top bioinformatics and computational biology conferences that is held around this time every year. Imagine coordinating with them to have the conference back-to-back with perhaps a day of overlapped joint sessions. The exchange of ideas could be really important. You could imagine other such partnerships with conference in engineering, art, etc. Jason

Comment on Sketch of a 2-player connection game by Sketch of an “artificial scientist” project | Pragmatic Genetic Programming

Fri, 29 Jun 2012 16:04:01 +0000

[…] seem to like mathematical recreations. Just the other day I sketched a little game thing that somebody could pick up and run. If it sits around long […]