Bayesian estimates, stable with respect to the choice of the loss function
Matematičeskie zametki, Tome 23 (1978) no. 2, pp. 327-334
Voir la notice de l'article provenant de la source Math-Net.Ru
A family of distributions is defined for which the generalized Bayesian estimate of a real parameter $\theta$, constructed according to the repeated choice, does not depend on the choice of the even convex loss function from a sufficiently wide class. It is shown that these families are a subclass of the exponential families with a sufficient statistic for the parameter $\theta$ of rank two.
@article{MZM_1978_23_2_a16,
author = {L. B. Klebanov},
title = {Bayesian estimates, stable with respect to the choice of the loss function},
journal = {Matemati\v{c}eskie zametki},
pages = {327--334},
publisher = {mathdoc},
volume = {23},
number = {2},
year = {1978},
language = {ru},
url = {http://geodesic.mathdoc.fr/item/MZM_1978_23_2_a16/}
}
L. B. Klebanov. Bayesian estimates, stable with respect to the choice of the loss function. Matematičeskie zametki, Tome 23 (1978) no. 2, pp. 327-334. http://geodesic.mathdoc.fr/item/MZM_1978_23_2_a16/