Automated Discovery of Numerical Approximation Formulae Via Genetic Programming

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

@MastersThesis{streeter:masters,
  author =       "Matthew J. Streeter",
  title =        "Automated Discovery of Numerical Approximation
                 Formulae Via Genetic Programming",
  school =       "Computer Science",
  year =         "2001",
  address =      "Worcester Polytechnic Institute, MA, USA",
  month =        may,
  keywords =     "genetic algorithms, genetic programming,
                 approximations, machine learning, artificial
                 intelligence",
  URL =          "http://www.wpi.edu/Pubs/ETD/Available/etd-0426101-231555/unrestricted/streeter.pdf",
  size =         "102 pages",
  abstract =     "This thesis describes the use of genetic programming
                 to automate the discovery of numerical approximation
                 formulae. Results are presented involving rediscovery
                 of known approximations for Harmonic numbers and
                 discovery of rational polynomial approximations for
                 functions of one or more variables, the latter of which
                 are compared to Pade approximations obtained through a
                 symbolic mathematics package. For functions of a single
                 variable, it is shown that evolved solutions can be
                 considered superior to Pade approximations, which
                 represent a powerful technique from numerical analysis,
                 given certain tradeoffs between approximation cost and
                 accuracy, while for functions of more than one
                 variable, we are able to evolve rational polynomial
                 approximations where no Pade approximation can be
                 computed. Furthermore, it is shown that evolved
                 approximations can be iteratively improved through the
                 evolution of approximations to their error function.
                 Based on these results, we consider genetic programming
                 to be a powerful and effective technique for the
                 automated discovery of numerical approximation
                 formulae.",
  notes =        "etd-0426101-231555",
}

Genetic Programming entries for Matthew J Streeter

Citations