Loading ...
Sorry, an error occurred while loading the content.
 

Re: about the innovation number [1 Attachment]

Expand Messages
  • Nikolai
    ... thanks Ken. i will try Modular NEAT too however, and compare it to Hyper NEAT after that, i think on a big cluster working on small problem modular NEAT
    Message 1 of 15 , Jul 1, 2009
      --- In neat@yahoogroups.com, "Kenneth Stanley" <kstanley@...> wrote:
      >
      > Nikolai, the idea of having a gene encapsulate a number of components is similar to the idea in Modular NEAT from Joeseph Reisinger a few years ago:
      >
      > http://nn.cs.utexas.edu/keyword?reisinger:gecco04
      >

      thanks Ken.
      i will try Modular NEAT too however, and compare it to Hyper NEAT after that, i think on a big cluster working on small problem modular NEAT could give more quality.
    • Robin Ledig
      Hello Nikolai, I did some extensive testing with the modular NEAT code, it is strongly bugged. I tried to fix them but gave up after 6 Weeks. Also I found some
      Message 2 of 15 , Jul 1, 2009
        Hello Nikolai,

        I did some extensive testing with the modular NEAT code, it is strongly bugged.
        I tried to fix them but gave up after 6 Weeks. Also I found some problems where the modular NEAT approach
        wont be better than NEAT because the results of two or more modules are not used in further calculation.
        I hope to build a hierarchical hyperNEAT algorithm with a dynamic modul approach which fixes those problems
        in my diploma thesis. It will also implement a dynamic substrat configuration, so you might want to wait for that.

        Robin

        --- In neat@yahoogroups. com, "Kenneth Stanley" <kstanley@.. .> wrote:
        >
        > Nikolai, the idea of having a gene encapsulate a number of components is similar to the idea in Modular NEAT from Joeseph Reisinger a few years ago:
        >
        > http://nn.cs. utexas.edu/ keyword?reisinge r:gecco04
        >

        thanks Ken.
        i will try Modular NEAT too however, and compare it to Hyper NEAT after that, i think on a big cluster working on small problem modular NEAT could give more quality.


      • Kenneth Stanley
        Robin, just curious, what do you mean by dynamic substrate configuration? How will that work? ken
        Message 3 of 15 , Jul 1, 2009
          Robin, just curious, what do you mean by "dynamic substrate configuration?" How will that work?

          ken

          --- In neat@yahoogroups.com, Robin Ledig <mai02igw@...> wrote:
          >
          > Hello Nikolai,
          >
          > I did some extensive testing with the modular NEAT code, it is strongly
          > bugged.
          > I tried to fix them but gave up after 6 Weeks. Also I found some
          > problems where the modular NEAT approach
          > wont be better than NEAT because the results of two or more modules are
          > not used in further calculation.
          > I hope to build a hierarchical hyperNEAT algorithm with a dynamic modul
          > approach which fixes those problems
          > in my diploma thesis. It will also implement a dynamic substrat
          > configuration, so you might want to wait for that.
          >
          > Robin
          >
          > >
          > >
          > > --- In neat@yahoogroups.com <mailto:neat%40yahoogroups.com>, "Kenneth
          > > Stanley" <kstanley@> wrote:
          > > >
          > > > Nikolai, the idea of having a gene encapsulate a number of
          > > components is similar to the idea in Modular NEAT from Joeseph
          > > Reisinger a few years ago:
          > > >
          > > > http://nn.cs.utexas.edu/keyword?reisinger:gecco04
          > > <http://nn.cs.utexas.edu/keyword?reisinger:gecco04>
          > > >
          > >
          > > thanks Ken.
          > > i will try Modular NEAT too however, and compare it to Hyper NEAT
          > > after that, i think on a big cluster working on small problem modular
          > > NEAT could give more quality.
          > >
          > >
          >
        Your message has been successfully submitted and would be delivered to recipients shortly.