Automatically Choosing the Number of Fitness Cases: The Rational Allocation of Trials

Created by W.Langdon from gp-bibliography.bib Revision:1.4216

  author =       "Astro Teller and David Andre",
  title =        "Automatically Choosing the Number of Fitness Cases:
                 The Rational Allocation of Trials",
  booktitle =    "Genetic Programming 1997: Proceedings of the Second
                 Annual Conference",
  editor =       "John R. Koza and Kalyanmoy Deb and Marco Dorigo and 
                 David B. Fogel and Max Garzon and Hitoshi Iba and 
                 Rick L. Riolo",
  year =         "1997",
  month =        "13-16 " # jul,
  keywords =     "genetic algorithms, genetic programming",
  pages =        "321--328",
  address =      "Stanford University, CA, USA",
  publisher_address = "San Francisco, CA, USA",
  publisher =    "Morgan Kaufmann",
  URL =          "",
  size =         "8 pages",
  abstract =     "For many problems to which genetic programming has
                 been applied, choosing the number of fitness cases with
                 which to evaluate the individuals is a crucial
                 decision. If too few fitness cases are used,
                 overfitting may occur, and the measured fitness of an
                 individual may not be representative of its true
                 fitness. On the other hand, if too many fitness cases
                 are used, a great deal of computer time can be wasted.
                 This paper presents a method for the Rational
                 Allocation of Trials (RAT) that dynamically allocates a
                 boundedly optimal number of fitness cases for each
                 individual. RAT allocates individuals to tournaments
                 prior to their evaluation, and then, borrowing from
                 previous work in model selection, allocates trials
                 (fitness cases) only to those individuals for whom the
                 cost of evaluating another fitness case is outweighed
                 by the expected utility that the new information will
                 provide. For most evolutionary computation approaches,
                 including genetic programming, and for most problems,
                 the RAT algorithm will provide significant time savings
                 at minimal additional system complexity.",
  notes =        "GP-97",

Genetic Programming entries for Astro Teller David Andre