Publication detail

The Formal Stochastic Framework for Comparison of Genetic Algorithms

POPELA, P., ROUPEC, J., OŠMERA, P., MATOUŠEK, R.

Original Title

The Formal Stochastic Framework for Comparison of Genetic Algorithms

English Title

The Formal Stochastic Framework for Comparison of Genetic Algorithms

Type

conference paper

Language

en

Original Abstract

The paper purpose is to discuss the comparison of GAs. The iterations are considered as random element realizations for GAs formally defined. The quantification of algorithm capabilities inspires the use of statistical methods. The significant difference between various setups is detected by statistical tests for the test problem.

English abstract

The paper purpose is to discuss the comparison of GAs. The iterations are considered as random element realizations for GAs formally defined. The quantification of algorithm capabilities inspires the use of statistical methods. The significant difference between various setups is detected by statistical tests for the test problem.

Keywords

genetic algorithms, statical tests, random elements

RIV year

2002

Released

12.05.2002

Publisher

IEEE

Location

Honolulu, Hawaii, USA

ISBN

0-7803-7281-6

Book

The 2002 IEEE World Congress on Computational Intelligence

Pages from

576

Pages to

581

Pages count

6

BibTex


@inproceedings{BUT5795,
  author="Pavel {Popela} and Jan {Roupec} and Pavel {Ošmera} and Radomil {Matoušek}",
  title="The Formal Stochastic Framework for Comparison of Genetic Algorithms",
  annote="The paper purpose is to discuss the comparison of GAs. The iterations are considered as random element realizations for GAs formally defined. The quantification of algorithm capabilities inspires the use of statistical methods. The significant difference between various setups is detected by statistical tests for the test problem.",
  address="IEEE",
  booktitle="The 2002 IEEE World Congress on Computational Intelligence",
  chapter="5795",
  institution="IEEE",
  year="2002",
  month="may",
  pages="576",
  publisher="IEEE",
  type="conference paper"
}