Detail publikace

Evolution of Iterative Formulas Using Cartesian Genetic Programming

MINAŘÍK, M. SEKANINA, L.

Originální název

Evolution of Iterative Formulas Using Cartesian Genetic Programming

Anglický název

Evolution of Iterative Formulas Using Cartesian Genetic Programming

Jazyk

en

Originální abstrakt

Many functions such as division or square root are implemented in hardware using iterative algorithms. We propose a genetic programming-based method to automatically design simple iterative algorithms from elementary functions. In particular, we demonstrated that Cartesian Genetic Programming can evolve various iterative formulas for tasks such as division or determining the greatest common divisor using a reasonable computational effort.

Anglický abstrakt

Many functions such as division or square root are implemented in hardware using iterative algorithms. We propose a genetic programming-based method to automatically design simple iterative algorithms from elementary functions. In particular, we demonstrated that Cartesian Genetic Programming can evolve various iterative formulas for tasks such as division or determining the greatest common divisor using a reasonable computational effort.

Dokumenty

BibTex 


@article{BUT76353,
  author="Miloš {Minařík} and Lukáš {Sekanina}",
  title="Evolution of Iterative Formulas Using Cartesian Genetic Programming",
  annote="Many functions such as division or square root are implemented in hardware using
iterative algorithms. We propose a genetic programming-based method to
automatically design simple iterative algorithms from elementary functions. In
particular, we demonstrated that Cartesian Genetic Programming can evolve various
iterative formulas for tasks such as division or determining the greatest common
divisor using a reasonable computational effort.",
  address="NEUVEDEN",
  chapter="76353",
  edition="NEUVEDEN",
  howpublished="print",
  institution="NEUVEDEN",
  number="6881",
  volume="2011",
  year="2011",
  month="september",
  pages="11--20",
  publisher="NEUVEDEN",
  type="journal article - other"
}