Surrogates and (mono-objective) optimization: a long-term relationship

Abstract : Surrogates have become an important part of recent optimization methods because they allow to deal with expensive functions. This lecture, that was given as part of the SAMCO (Surrogate- Assisted Multi­Criteria Optimization) workshop at the Lorentz Center in February 2016, summarizes the different ways in which surrogates can contribute to optimization algorithms. The course is introduced with two naive (not converging) algorithms, stressing the need for controls that ensure a useful contribution of the surrogate to the optimization. Then, a first class of solutions, based on trust regions, is presented. Next, we discuss implementations where surrogates support stochastic optimization algorithms. The third family of approaches is based on surrogates that provide, in addition to a prediction of the objective function, an estimate of the prediction error: this concerns mainly kriging surrogates. The last part of the course touches upon the topic of ensembles of surrogates.
Type de document :
Cours
École thématique. Lorentz Center, Leiden, Netherlands. 2016
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/cel-01281650
Contributeur : Le Riche Rodolphe <>
Soumis le : mercredi 2 mars 2016 - 14:58:42
Dernière modification le : jeudi 6 décembre 2018 - 15:03:54
Document(s) archivé(s) le : vendredi 3 juin 2016 - 11:22:05

Identifiants

  • HAL Id : cel-01281650, version 1

Citation

Rodolphe Le Riche. Surrogates and (mono-objective) optimization: a long-term relationship. École thématique. Lorentz Center, Leiden, Netherlands. 2016. 〈cel-01281650〉

Partager

Métriques

Consultations de la notice

618

Téléchargements de fichiers

548