Approximate optimal designs for multivariate polynomial regression

Abstract : We introduce a new approach aiming at computing approximate optimal designs for multivariate polynomial regressions on compact (semialgebraic) design spaces. We use the moment-sum-of-squares hierarchy of semidefinite programming problems to solve numerically the approximate optimal design problem. The geometry of the design is recovered via semidefinite programming duality theory. This article shows that the hierarchy converges to the approximate optimal design as the order of the hierarchy increases. Furthermore, we provide a dual certificate ensuring finite convergence of the hierarchy and showing that the approximate optimal design can be computed numerically with our method. As a byproduct, we revisit the equivalence theorem of the experimental design theory: it is linked to the Christoffel polynomial and it characterizes finite convergence of the moment-sum-of-square hierarchies.
Type de document :
Article dans une revue
Annals of Statistics, Institute of Mathematical Statistics, 2019, 47 (1), pp.127-155. 〈10.1214/18-AOS1683〉
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-01980595
Contributeur : Fabrice Gamboa <>
Soumis le : lundi 14 janvier 2019 - 15:23:12
Dernière modification le : jeudi 7 février 2019 - 16:03:33

Lien texte intégral

Identifiants

Citation

Yohann De Castro, Fabrice Gamboa, Didier Henrion, Roxana Hess, Jean-Bernard Lasserre. Approximate optimal designs for multivariate polynomial regression. Annals of Statistics, Institute of Mathematical Statistics, 2019, 47 (1), pp.127-155. 〈10.1214/18-AOS1683〉. 〈hal-01980595〉

Partager

Métriques

Consultations de la notice

28