On ANOVA decompositions of kernels and Gaussian random field paths

Abstract : The FANOVA (or "Sobol'-Hoeffding") decomposition of multivariate functions has been used for high-dimensional model representation and global sensitivity analysis. When the objective function f has no simple analytic form and is costly to evaluate, a practical limitation is that computing FANOVA terms may be unaffordable due to numerical integration costs. Several approximate approaches relying on random field models have been proposed to alleviate these costs, where f is substituted by a (kriging) predictor or by conditional simulations. In the present work, we focus on FANOVA decompositions of Gaussian random field sample paths, and we notably introduce an associated kernel decomposition (into 2^{2d} terms) called KANOVA. An interpretation in terms of tensor product projections is obtained, and it is shown that projected kernels control both the sparsity of Gaussian random field sample paths and the dependence structure between FANOVA effects. Applications on simulated data show the relevance of the approach for designing new classes of covariance kernels dedicated to high-dimensional kriging.
Liste complète des métadonnées

Littérature citée [36 références]  Voir  Masquer  Télécharger

https://hal.archives-ouvertes.fr/hal-01066503
Contributeur : David Ginsbourger <>
Soumis le : jeudi 2 octobre 2014 - 13:11:25
Dernière modification le : mardi 23 octobre 2018 - 14:36:09
Document(s) archivé(s) le : samedi 3 janvier 2015 - 10:56:12

Fichiers

Ginsbourger_etal.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01066503, version 2
  • ARXIV : 1409.6008

Citation

David Ginsbourger, Olivier Roustant, Dominic Schuhmacher, Nicolas Durrande, Nicolas Lenz. On ANOVA decompositions of kernels and Gaussian random field paths. 2014. 〈hal-01066503v2〉

Partager

Métriques

Consultations de la notice

770

Téléchargements de fichiers

251