BloJJ

Adventures of a multidimensional freak

This is Juan Julián Merelo Guervós English-language blog. He teaches computer science at the University of Granada, in southern Spain. Come back here to read about politics, technology, with a new twist

Latest comments

  • More Info en Compute the number of cites in Google Citations
  • uali20 en Äppärätti in Super Sad True Love Story
  • uali20 en Cloning part of a local repository
  • robert en About conference poster design and defense
  • Jackson en About conference poster design and defense
  • Nano Towels where to buy en Ants and roaches at Deneubourg's lab
  • besiemrgns01 en Riddles in Kafka on the shore
  • Jaxon en Riddles in Kafka on the shore
  • wood grilling planks en Riddles in Kafka on the shore
  • robert en Riddles in Kafka on the shore
  • Blogs Out There

    Nelson Minar's Blog
    Jeremy Zawodny's Blog
    Kottke
    Complexes, Carlos Gershenson's blog
    IlliGAL, Genetic Algorithms blog
    Blogging in the wind, Víctor R. Ruiz's blog


    Atalaya, my Spanish language blog
    Geneura@Wordpress, our research group's blog.
    My home page

    Old stories


    Creative Commons License
    This work is licensed under a Creative Commons License.
    Blogalia

    Stats
    Inicio > Historias > Paper on dissortative mating genetic algorithms uploaded to ArXiV

    Paper on dissortative mating genetic algorithms uploaded to ArXiV

    As our mate Carlos tells in our research team blog, we have uploaded our paper Using Dissortative Mating Genetic Algorithms to Track the Extrema of Dynamic Deceptive Functions, which will eventually become a part of his PhD thesis.
    ADMGA (adaptive-threshold dissortative mating GAs) is a nifty idea: try and preserve diversity by making individuals in the population only mate with those that are different enough. Diversity is always important in GAs, but even more so in problems where memory is a bonus, like dynamic optimization problems.
    Results obtained have been quite good, even more so with the hardest instances. So, good luck with your PhD, Carlos.
    Etiquetas: , , ,

    2009-04-22 13:35 | 6 Comment(s) | Filed in Research

    Referencias (TrackBacks)

    URL de trackback de esta historia http://blojj.blogalia.com//trackbacks/62788

    Comentarios

    1
    De: Marcelo Fecha: 2009-04-23 21:42

    Hi, JJ!

    Interesting idea that one.

    It reminded me of a diversity-guided evolutionary algorithm (a GA) I implemented some years ago -- I think during 2005. Rasmus Ursem, the author, had applied a "similar" concept to deal with the optimization process:

    "The DGEA applies diversity-decreasing operators (selection and recombination) as long as the diversity is above a certain threshold dlow . When the diversity drops below dlow the DGEA switches to diversity-increasing operators (mutation) until a diversity of dhigh is reached. Hence, phases with exploration and phases with exploitation will occur (see Fig. 2). Theoretically, the DGEA should be able to escape local optima because the operators will force higher diversity regardless of fitness."

    Of course the user *must* use elitism, otherwise the aforementioned algorithm will never converge to an optimum and would perform a simple random walk. That DGEA is very CPU intensive and, if not well implemented (mainly the distance-to-average measure), tend to call so many times the evolutionary operators, without giving a satisfatory result. At least was that what I got after implementing it. But I do not know if I implemented something wrong, since I only had on my hands the paper and Ursem's email to ask him some questions. It was fun to implement that algorithm! :)

    I found a misspelling in Carlos' paper:

    "Besides *de* above-referred techniques[...]."

    I think he meant "the", not the Portuguese preposition "de".

    Hey, JJ, could you send me, please, a pdf copy of the following paper?

    "Comparing evolutionary hybrid systems for design and optimization of multilayer perceptron structure along training parameters."

    I would be very grateful if I could read it! The email is that one in the "Correo-e" text box.

    Thank you in advance!

    Até!

    Marcelo



    2
    De: JJ Fecha: 2009-04-23 22:11

    Sure... and thanks for the corrections. There are probably some more errors; it's a draft, after all.
    I think I know Ursem's work; I'll have to check the latest references. We have been working on lots of things related to diversity lately.



    3
    De: Marcelo Fecha: 2009-04-24 20:29

    Hi, JJ!

    Diversity research in evolutionary computation is an amazing subject for investigation. I think it is one of the roads EC researchers should follow to overcome and improve the classic elitist SGA. I hope your group may find a way of avoiding large populations when it comes to genetic algorithms. I have seen some problems -- a la trap functions -- in which the user should set a big population size, such as 1000 or 2000 for a 50 or 100 dimension problem.

    JJ, I do not intend to hurry you, but did you send me the aforementioned paper above? If you are too busy, you can send it to me in a more suitable moment.

    Congratulations for the research! :)

    Até!

    Marcelo



    4
    De: JJ Fecha: 2009-04-24 20:51

    Damn, I knew I had to send a paper to somebody today, and I looked up my email, but I didn't find it... please send me an email and I will send it to you next Monday (or before)



    5
    De: Marcelo Fecha: 2009-04-29 18:10

    Hi, JJ!

    Thank you very much for sending me the paper I had requested. :)

    I shall read it next weekend and, if I have any comment to say, I'll send you an e-mail. :)

    Gracias!

    Marcelo



    6
    De: JJ Fecha: 2009-05-04 09:29

    No problem. Enjoy it!



    Nombre
    Correo-e
    URL
    Dirección IP: 54.196.127.107 (ac72f700ee)
    Comentario
    ¿Cuánto es: mil + uno?

    © 2002 - 2008 jmerelo
    Powered by Blogalia