GP.Lab

about

A Genetic Programming workbench to analyze and automatically solve problems with this machine learning technique. To learn more about Genetic Algorithms and Genetic Programming in particular have a read at Wikipedia.

GP.Lab uses the ESS-GP kernel to automatically solve various problems. The list of currently implemented modules includes:

  • Mathematical function regression
  • Santa Fe Trail problem
  • RoboRail
  • Collector (temp. disabled)


The current demo is a very, very early preview but it might give you the idea.
We hope to provide regular updates and incorporate more modules in the near future!

Our YouTube channel www.youtube.com/virtualspecies/ features some short videos for various GP.Lab modules.

Any kind of feedback is welcome.

news

Current efforts are focussing on two areas:

  • Statistics

    The various statistic counters already gathered throughout the kernel are being collated in a new ESS-GP module and vastly extended to provide a much deeper insight into the population and the progress of the simulated evolution.
    This will include real-time graphs in GP.Lab of all relevant counters and export of any counter to GNUPLOT.
  • Refactoring the ESS-GP kernel

    The ESS-GP kernel is currently being refactored to improve adaptability for future modules and to prepare it for a generic task mechanism to support concurrent computation of statistics, population evaluation and more not bound to any particular architecture (i.e. threads, tasks, cloud, etc.).

screenshots

donate

Feel free to make a small donation if you like the program and would like to support the work on this freeware application.