Wednesday, September 07, 2005

 

Common criticisms of GAs

are usually ill-informed. The blogosphere provides regular examples of this sort of drivel, for example here and here. The first is a rant about how general purpose solvers are doomed and the second complains how machine invention is something of a Rube Goldberg machine.

If blind optimization is as inadequate as NFLers insist, how did nature evolve the incredible complexity surrounding us? And if genetic algorithms and genetic programming are so weird, how come these procedures are regularly infringing on patents of human inventors and creating new patentable gizmos?

Many of these complainants seem stuck in a time warp of GA research as it was ten years ago or so (20?). Please wake up and smell the coffee. The field of genetic and evolutionary computation is roasting hot beans, grinding them just right, and brewing a lovely cup of Joe. Instead of griping about some perceived difficulty within the bowels of our percolator, you might want to take a sip, and learn why this field, not only continues to survive, but insists on thriving, flourishing, and breathing life into research across the spectrum of human endeavor.

Comments:
Be careful Prof. Goldberg, your experience in the field does not make you immune to producing the same 'drivel' that you so openly criticise here. The blogs you so easily dismiss as 'drivel' produce some interesting discussion on the topic of GA's producing useful/useless? solution features, and are well worth a read (at least the second link).
NFL and those who 'subscribe' to it are not nescesarily having a go at GA's or any other search algorithms for that matter, NFL simply points out that there are 'theoretical' limitations imposed on any algorithmic process when no prior information is available. Just because a person chooses to think about and discuss NFL does not make them a GA hater.
 
Hallo!! :D

Once in this blog I had said that the NFL and DeJong's ideas about GAFO, both are not sufficient reasons to stop us, the Evolutionary Algorithmists, to research and try to shed light on the questions that EA's offer us.

Here in Brazil, in a considerable number of Universities, Soft Computing approaches(Neural Nets, Fuzzy Logic and Evolutionary Algorithms) are not so well considered among a lot of Professors, Researchers and Students. Here where I study there is even a Professor who says that Genetic Algorithms are for fags/gays and he, the Professor, prefers to use classical methods(Operations Research, Markovian Models). I had last semester a course in Performance Analysis of Systems, basically it was a little introduction into the Markovian World, and the guy who was teaching it was being supervised in his Phd Thesis by the Professor who thinks that EA's are for fags. That guy always he had a chance to provoke me, he did it saying a lot of drivels ("The last chapter of a Neural Network Book is the first chapter of my Phd Thesis!" - Because he knows that my supervisor has research in the Neural Net field) and he even does not know what is NFL or GAFO. That guy said more: "I must be excellent in my field (Markovian Models) and, at same time, I must have a critic view about the other fields, like Genetic Algorithms!". I only think very strange that the criticizers of EAs never implemented an EA to say anything about and when they want to criticize, for example Genetic Algorithms, they critics only the SGA(that solves nothing) and say the same thing about: Binary Codification, Blind Search, the EA field is so brand new and so on. But a thing that is very very very funny is that that guy knows where is the room of my supervisor, he knows who is my supervisor, however, that guy never went to the room of my supervisor to talk about Soft Computing and to say that the first chapter of his Phd Thesis is the last chapter of a Neural Net book. And I ask you: Why didn't the guy go to the room of my supervisor to talk about Soft Computing ? Very simple: the guy knows nothing about the field and says a lot of drivels because he fears to lose his reputation among his lab mates, he needs to show that his field is the best.

Well, I do not know if that kind of view is bad for the Evolutionary Algorithmists, sometimes I guess that those persons, who are very ill-informed about EA, are good thing for us, because we know the power that we have and what we can build with it.

Até Mais!!

Marcelo (a.k.a Nosophorus)
 
Hallo!! :D

Well, I always wanted to ask this and I'm sorry if I'm being impolite, Dr Goldberg, but could you, Dr Goldberg, tell us some of the drivels that you heard about Evolutionary Computation along your Genetic Algorithmist life ? I only know two: "The Engineer with Odd credentials" and that other about the "Journal of Hydraulic Engineering".

Até Mais!!(Until Later!!)

Marcelo(a.k.a Nosophorus)
 
Hallo!!

Well, I'm sorry to had asked for some drivels that Dr. Goldberg have heard during his career. Sorry.

Até Mais!! :D

Marcelo
 
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?