Statistical multi-comparison of evolutionary algorithms

Résultats de recherche: Chapitre dans un livre, rapport, actes de conférenceParticipation à un ouvrage collectif lié à un colloque ou une conférenceRevue par des pairs

3 Citations (Scopus)

Résumé

Performance benchmarking is an essential task in computational intelligence research. Evolutionary optimization algorithms are stochastic processes and to obtain significant results proper statistical tools must be used. This paper presents a step-by-step nonparametric comparison procedure able to assess the relative performance of several singleobjective evolutionary algorithms. This comparison procedure is based on a simple ranking scheme and is statistically relevant under a controlled risk of error. A useful feature of the procedure is the ability to visualize the dynamical behavior of the algorithms along with the confidence intervals of their performance. It also has the advantage of eliminating the use of arbitrary weighting coefficients when several comparisons criteria are involved.

langue originaleAnglais
titreBioinspired Optimization Methods and their Applications - Proceedings of the 3rd International Conference on Bioinspired Optimization Methods and their Applications, BIOMA 2008
Pages71-80
Nombre de pages10
étatPublié - 2008
Evénement3rd International Conference on Bioinspired Optimization Methods and their Applications, BIOMA 2008 - Ljubljana, Slovénie
Durée: 13 oct. 200814 oct. 2008

Série de publications

NomBioinspired Optimization Methods and their Applications - Proceedings of the 3rd International Conference on Bioinspired Optimization Methods and their Applications, BIOMA 2008

Conférence

Conférence3rd International Conference on Bioinspired Optimization Methods and their Applications, BIOMA 2008
Pays/TerritoireSlovénie
La villeLjubljana
période13/10/0814/10/08

Empreinte digitale

Voici les principaux termes ou expressions associés à « Statistical multi-comparison of evolutionary algorithms ». Ces libellés thématiques sont générés à partir du titre et du résumé de la publication. Ensemble, ils forment une empreinte digitale unique.

Contient cette citation