A general framework for statistical performance comparison of evolutionary computation algorithms

  • Authors:

  • David Shilane;Jarno Martikainen;Sandrine Dudoit;Seppo J. Ovaska

  • Affiliations:

  • Division of Biostatistics, School of Public Health, University of California, Berkeley, CA, USA;Faculty of Electronics, Communications, and Automation, Helsinki University of Technology, Otakaari 5 A, P.O. Box 3000, 02015 Espoo, Finland;Division of Biostatistics, School of Public Health, University of California, Berkeley, CA, USA;Faculty of Electronics, Communications, and Automation, Helsinki University of Technology, Otakaari 5 A, P.O. Box 3000, 02015 Espoo, Finland

  • Venue:
  • Information Sciences: an International Journal
  • Year:
  • 2008

Quantified Score

Hi-index 0.08

Visualization

Abstract

This paper proposes a statistical methodology for comparing the performance of evolutionary computation algorithms. A twofold sampling scheme for collecting performance data is introduced, and these data are analyzed using bootstrap-based multiple hypothesis testing procedures. The proposed method is sufficiently flexible to allow the researcher to choose how performance is measured, does not rely upon distributional assumptions, and can be extended to analyze many other randomized numeric optimization routines. As a result, this approach offers a convenient, flexible, and reliable technique for comparing algorithms in a wide variety of applications.