Guaranteed characterization of exact non-asymptotic confidence regions

Abstract : In parameter estimation, it is often desirable to supplement the estimates with an assessment of their quality. A new family of methods proposed by Campi et al. for this purpose is particularly attractive, as it makes it possible to obtain exact, non-asymptotic con dence regions under mild assumptions on the noise distribution. A bottleneck of this approach, however, is the numerical characterization of these con dence regions. So far, it has been carried out by gridding, which provides no guarantee as to its results and is only applicable to low dimensional spaces. This paper shows how interval analysis can contribute to removing this bottleneck.
Liste complète des métadonnées

Littérature citée [10 références]  Voir  Masquer  Télécharger

https://hal-supelec.archives-ouvertes.fr/hal-00879937
Contributeur : Michel Kieffer <>
Soumis le : mardi 5 novembre 2013 - 10:18:21
Dernière modification le : mercredi 26 septembre 2018 - 01:15:51
Document(s) archivé(s) le : jeudi 6 février 2014 - 04:35:25

Fichier

Automatica_Revised_v4.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-00879937, version 1

Citation

Michel Kieffer, Eric Walter. Guaranteed characterization of exact non-asymptotic confidence regions. accepté à Automatica. 2013. 〈hal-00879937〉

Partager

Métriques

Consultations de la notice

556

Téléchargements de fichiers

133