A class of Bivariate SURE estimators in heteroscedastic hierarchical normal models

In this paper, we first propose a class of bivariate shrinkage estimators based on Steins unbiased estimate of risk (SURE). Then, we study the effect of correlation coefficients on their performance. Moreover, under some mild assumptions on the model correlations, we set up the optimal asymptotic pr...

Full description

Saved in:
Bibliographic Details
Published in:Journal of statistical theory and applications 2018-06, Vol.17 (2), p.324-339
Main Author: Ghoreishi, S. K.
Format: Article
Language:English
Subjects:
Asymptotic properties
Asymptotic univariate shrinkage estimators
Bivariate analysis
Correlation coefficients
Estimators
Heteroscedasticity
Hierarchical models
Multivariate shrinkage estimator
Research Article
Stein’s unbiased risk estimators
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this paper, we first propose a class of bivariate shrinkage estimators based on Steins unbiased estimate of risk (SURE). Then, we study the effect of correlation coefficients on their performance. Moreover, under some mild assumptions on the model correlations, we set up the optimal asymptotic properties of our estimates when the number of vector means to be estimated grows . Furthermore, we carry out a simulation study to compare how various bivariate competing shrinkage estimators perform and analyze a real data set.
ISSN:1538-7887
2214-1766
1538-7887
DOI:10.2991/jsta.2018.17.2.11