RT info:eu-repo/semantics/article
T1 A parametric registration model for warped distributions with Wasserstein’s distance.
A1 Agulló Antolín, Marina
A1 Cuesta Albertos, Juan Antonio
A1 Lescornel, Hélène
A1 Loubes, Jean-Michel
K1 Wasserstein distance
AB We consider a parametric deformation model for distributions. More precisely, we assumewe observe J samples of random variables which are warped from an unknown distributiontemplate. We tackle in this paper the problem of estimating the individual deformationparameters. For this, we construct a registering criterion based on the Wasserstein distanceto quantify the alignment of the distributions. We prove consistency of the empiricalestimators.
PB Elsevier
SN 0047-259X
YR 2015
FD 2015
LK http://uvadoc.uva.es/handle/10324/21813
UL http://uvadoc.uva.es/handle/10324/21813
LA eng
NO Journal of Multivariate Analysis, 2015, vol. 135, p. 117-130
NO Producción Científica
DS UVaDOC
RD 28-nov-2024