Creating Objects Using Genetic Programming Techniques

Created by W.Langdon from gp-bibliography.bib Revision:1.4420

  author =       "Yandu Oppacher and Franz Oppacher and Dwight Deugo",
  title =        "Creating Objects Using Genetic Programming
  booktitle =    "10th ACIS International Conference on Software
                 Engineering, Artificial Intelligences, Networking and
                 Parallel/Distributed Computing, SNPD '09",
  year =         "2009",
  month =        "27-29 " # may,
  editors =      "Haeng-Kon Kim and Roger Lee",
  address =      "Daegu, Korea",
  pages =        "455--461",
  keywords =     "genetic algorithms, genetic programming, grammatical
                 evolution, SBSE, JUnit test suite, Java program,
                 evolutionary computation, grammar-based genetic
                 programming, object oriented programming,
                 object-oriented encapsulation, Java, object-oriented
  DOI =          "doi:10.1109/SNPD.2009.82",
  isbn13 =       "978-0-7695-3642-2",
  abstract =     "Grammar-based object oriented genetic programming
                 leverages the object oriented programming paradigm in
                 evolutionary computation. The difference between our
                 approach and others is that the resulting solution is a
                 syntactically correct and executable Java program.
                 Candidate solutions are created and tested against a
                 JUnit test suite. Evolved solutions incorporate useful
                 object oriented patterns. To create and compile the
                 Java classes, a new framework for grammar-based genetic
                 programming was created. The framework is entirely
                 written in Java. By externally compiling and then
                 loading the classes into the framework for testing,
                 true object-oriented encapsulation is achieved. For the
                 first time state and behavior are passed around as an
  notes =        "GPOOGP. Codon values in the range [0;10]. Builder
                 class. Visitor Pattern. Java 6 compiler (no
                 interpreter). Fitness = sum of |error|. subtree
                 preserving crossover. Symbolic regression, two-box,
                 bank account, reverse polish notation calculator. BNF
                 by hand (not given). p460 'reflection...allowed the
                 grammar to be independent of the problem domain'.
                 double-dispatching. stack. XML Skeleton provides
                 wrapper around evolution. Also known as

Genetic Programming entries for Yandu Oppacher Franz Oppacher Dwight Deugo