Robust estimation with latin hypercube sampling: a central limit theorem for Z-estimators
Résumé
Latin hypercube sampling (LHS) is a stratified sampling
method widely used in computer experiments. In this
work, we extend convergence results on the sample mean
with Latin hypercube sampling to the class of Z -estimators,
gathering all estimators that can be written as zeros of a
sample mean function. In particular, the asymptotic vari-
ance of this estimate is obtained. This asymptotic vari-
ance is shown to be lower using LHS than using classic
independent and identically distributed sampling. A Cen-
tral Limit theorem for Z -estimators under LHS is also
given.
Fichier principal
A_Central_Limit_Theorem_for_LHS_Z_estimators (1).pdf (555.27 Ko)
Télécharger le fichier
Origine | Fichiers produits par l'(les) auteur(s) |
---|