A representation for the Adaptive Generation of Simple Sequential Programs

Created by W.Langdon from gp-bibliography.bib Revision:1.4549

  author =       "Nichael Lynn Cramer",
  title =        "A representation for the Adaptive Generation of Simple
                 Sequential Programs",
  year =         "1985",
  booktitle =    "Proceedings of an International Conference on Genetic
                 Algorithms and the Applications",
  address =      "Carnegie-Mellon University, Pittsburgh, PA, USA",
  month =        "24-26 " # jul,
  editor =       "John J. Grefenstette",
  pages =        "183--187",
  size =         "5 pages",
  URL =          "http://www.cs.ucl.ac.uk/staff/W.Langdon/ftp/papers/icga1985/icga85_cramer.pdf",
  URL =          "http://www.sover.net/~nichael/nlc-publications/icga85/index.html",
  keywords =     "genetic algorithms, genetic programming, memory",
  abstract =     "An adaptive system for generating short sequential
                 computer functions is described. The created functions
                 are written in the simple {"}number-string{"} language
                 JB, and in TB, a modified version of JB with a
                 tree-like structure. These languages have the feature
                 that they can be used to represent well-formed, useful
                 computer programs while still being amenable to
                 suitably defined genetic operators. The system is used
                 to produce two-input, single-output multiplication
                 functions that are concise and well-defined. Future
                 work, dealing with extensions to more complicated
                 functions and generalizations of the techniques, is
                 also discussed.",
  notes =        "The earliest description of the tree-like
                 representation and operators for use in the application
                 of Genetic Algorithms to computer programs -

                 Evolves a multiplier, {"}72% more often than control
                 sample{"} {"}PL- not fully Turing Equivalent{"},
                 addition of :SET and :BLOCK lead to JB language (nb a
                 list of statements language). JB has problems with
                 crossover -> TB which is as JB but instead of calls to
                 other statements, these other statements are expanded
                 in the first yielding a tree shaped syntax. Crossover
                 operator changed to deal with sub trees! Both languages
                 contain small numbers of global integers. TB Mutation
                 restricted to frindges of tree, ie leaves or first
                 level functions. Inversion: crossover within same
                 program! Goldberg(1989, p 303) says {"}Cramer does not
                 present any results from the use of JB in any genetic
                 trials; however he abandoned these first efforts
                 because of some limited computational

                 Fitness based, to some extent, upon internals of
                 program. Limits on prog size via fitness. Forced
                 timeout Goldberg(1987) says timed out progs fitness was

                 =>Smith,S.F. IJCAI-83

                 Publisher not known, sponsored by USA Navy.",

Genetic Programming entries for Nichael Lynn Cramer