Detail publikace

Coevolution in Cartesian Genetic Programming

Originální název

Coevolution in Cartesian Genetic Programming

Anglický název

Coevolution in Cartesian Genetic Programming

Jazyk

en

Originální abstrakt

Cartesian genetic programming (CGP) is a branch of genetic programming which has been utilized in various applications. This paper proposes to introduce coevolution to CGP in order to accelerate the task of symbolic regression. In particular, fitness predictors which are small subsets of the training set are coevolved with CGP programs. It is shown using five symbolic regression problems that the (median) execution time can be reduced 2-5 times in comparison with the standard CGP.

Anglický abstrakt

Cartesian genetic programming (CGP) is a branch of genetic programming which has been utilized in various applications. This paper proposes to introduce coevolution to CGP in order to accelerate the task of symbolic regression. In particular, fitness predictors which are small subsets of the training set are coevolved with CGP programs. It is shown using five symbolic regression problems that the (median) execution time can be reduced 2-5 times in comparison with the standard CGP.

BibTex


@inproceedings{BUT91456,
  author="Michaela {Drahošová} and Lukáš {Sekanina}",
  title="Coevolution in Cartesian Genetic Programming",
  annote="Cartesian genetic programming (CGP) is a branch of genetic programming which has
been utilized in various applications. This paper proposes to introduce
coevolution to CGP in order to accelerate the task of symbolic regression. In
particular, fitness predictors which are small subsets of the training set are
coevolved with CGP programs. It is shown using five symbolic regression problems
that the (median) execution time can be reduced 2-5 times in comparison with the
standard CGP.",
  address="Springer Verlag",
  booktitle="Proc. of the 15th European Conference on Genetic Programming",
  chapter="91456",
  doi="10.1007/978-3-642-29139-5_16",
  edition="Lecture Notes in Computer Science",
  howpublished="print",
  institution="Springer Verlag",
  year="2012",
  month="march",
  pages="182--193",
  publisher="Springer Verlag",
  type="conference paper"
}