Tuesday, May 16, 2006
Live blogging NIGEL
- Stewart W. Wilson: "Can We Do Captchas?"
- David E. Goldberg: "Searle, Intentionality, and the Future of Classifier Systems"
- Dipankar Dasgupta: "Artificial Immune Systems in Anomaly Detection"
- Lashon Booker: "A Retrospective Look at Classifier System Research"
Stewart challenged us to go beyond "place/value functions" and work on problems that were more open ended. I challenged the field to work on creating "intentional machines" using some of Searle's ideas contained in Mind, Language & Society. Dipankar presented an overview of his work on artificial immune systems. Lashon is up next.
Gosh!! Only Genetic-like/-based and Weird Methods! This is very strange to a supercomputing center where is being held a Gathering on Evolutionary Learning. Do you plan to show serious Evolutionary Computation? Since you are Gathering on Evolutionary Learning, why are not there the other paradigms of EC, such as Evolutionary Programming, Evolution Strategies and Genetic Programming??
Artificial Immune Systems! Sometimes I guess if we _REALLY_ need new bio-inspired methods and what kinds of innovations those new methods can bring us. To be honest, I guess we do not need them, because those weird new methods only serve as a reason to give their author the honor to be the inventor/creator of something bio-inspired. It is interesting (weird??) that today there are some persons that want to be the inventor/creator of some bio-inspired method. I prefer to invest my time and money in well investigated methods, such as Evolution Strategies and Genetic Algorithms.
"Why are we doing this? Will LCS/GBML become mainstream research?"
I guess both can become mainstream research. The audience will only need to filter the lies and the truths. But I wonder if those methods could have some day rigorous analyses "a la" Evolution Strategies. I guess they won't! :)
When people mean Genetic based, they don't confine themselves to Genetic algorithms (in your narrow sense). At a higher level all genetic and evolutionary algorithms have many similarities and it does disservice to the GEC community to partition them and try to act as though when one uses "genetic based" they somehow are implying/suggesting "not any other evolutionary method-based".
Also, one should not be too eager to dismiss new ideas, especially when you don't know much about them. I can imagine that people had the same kind of reaction when many new ideas /discoveries, which are commonly accepted today.
I knew what I said, Dr Sastry. When I said, in my narrow sense (according you), Genetic Based I did not confined myself to Genetic Algorithms, but to all kinds of Genetic-like algorithms. I hope you can understand this, if not, I can make a drawing with sanskrit explanations. :-)
"At a higher level all genetic and evolutionary algorithms have many similarities and it does disservice to the GEC community to partition them and try to act as though when one uses 'genetic based' they somehow are implying/suggesting 'not any other evolutionary method-based'."
So, let's forget the term "GEC - Genetic And Evolutionary Algorithms" and use only "EC - Evolutionary Computation", since the last one has a more general meaning.
"Also, one should not be too eager to dismiss new ideas[...]"
Remember that I made a supposition about those new (and weird, in my opinion) bio-inspired methods. Only further I gave my thoughts about those new bio-inspiered methods.
"[...] especially when you don't know much about them."
I talked about AIS and not about LCS.
"I can imagine that people had the same kind of reaction when many new ideas /discoveries, which are commonly accepted today."
Yeah! I also imagine it every time! But I do not have any kind of fear of innovation, I only asked if we _REALLY_ need those new bio-inspired methods.
I suppose I am not the only person here with a narrow sense, since in your publications there are only Genetic Algorithms and/or Genetic-like methods. As if Evolutionary Computation was only Genetic Algorithms and/or Genetic-like methods.