A Higher-Order Function Approach to Evolve Recursive Programs

Created by W.Langdon from gp-bibliography.bib Revision:1.3872

@InCollection{yu:2005:recurse,
  author =       "Tina Yu",
  title =        "A Higher-Order Function Approach to Evolve Recursive
                 Programs",
  booktitle =    "Genetic Programming Theory and Practice {III}",
  year =         "2005",
  editor =       "Tina Yu and Rick L. Riolo and Bill Worzel",
  volume =       "9",
  series =       "Genetic Programming",
  chapter =      "7",
  pages =        "93--108",
  address =      "Ann Arbor",
  month =        "12-14 " # may,
  publisher =    "Springer",
  keywords =     "genetic algorithms, genetic programming, recursion,
                 Fibonacci sequence, strstr, PolyGP, type systems,
                 higher-order functions, recursion patterns, filter,
                 foldr, scanr, lambda abstraction, functional
                 programming languages, Haskell",
  ISBN =         "0-387-28110-X",
  URL =          "http://www.cs.mun.ca/~tinayu/Publications_files/gptp2005.pdf",
  DOI =          "doi:10.1007/0-387-28111-8_7",
  size =         "16 pages",
  abstract =     "We demonstrate a functional style recursion
                 implementation to evolve recursive programs. This
                 approach re-expresses a recursive program using a
                 non-recursive application of a higher-order function.
                 It divides a program recursion pattern into two parts:
                 the recursion code and the application of the code.
                 With the higher-order functions handling recursion code
                 application, GP effort becomes focused on the
                 generation of recursion code.

                 We employed this method to evolve two recursive
                 programs: strstr C library function and programs that
                 produce the Fibonacci sequence. In both cases, the
                 program space defined by higher-order functions are
                 very easy for GP to find a solution. We have learned
                 about higher-order function selection and fitness
                 assignment through this study. The next step will be to
                 test the approach on applications with open-ended
                 solutions, such as evolutionary design.",
  notes =        "part of \cite{yu:2005:GPTP} Published Jan 2006 after
                 the workshop",
}

Genetic Programming entries for Tina Yu

Citations