Empirical Bayes two-action problem for the continuous one-parameter exponential family with errors in variables

Rohana J. Karunamuni, Shunpu Zhang

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

In this paper we study the empirical Bayes linear-loss two-action problem for the continuous one-parameter exponential family when the observed data are contaminated (errors in variables). A new empirical Bayes testing rule is constructed, and its asymptotic optimality uniformly over a class of prior distributions is established. Uniform rates of convergence of the corresponding regret (excess risk), which depends on the type of the error distribution, are also obtained for two types of error distributions. Our results are compared with the 'pure' observed data results of the literature.

Original languageEnglish (US)
Pages (from-to)437-449
Number of pages13
JournalJournal of Statistical Planning and Inference
Volume113
Issue number2
DOIs
StatePublished - May 1 2003

Fingerprint

Errors in Variables
Empirical Bayes
Exponential Family
Asymptotic Optimality
Regret
Prior distribution
Excess
Rate of Convergence
Testing
Exponential family
Errors in variables
Class

Keywords

  • Asymptotically optimal
  • Empirical Bayes
  • Errors in variables
  • Rates of convergence

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty
  • Applied Mathematics

Cite this

Empirical Bayes two-action problem for the continuous one-parameter exponential family with errors in variables. / Karunamuni, Rohana J.; Zhang, Shunpu.

In: Journal of Statistical Planning and Inference, Vol. 113, No. 2, 01.05.2003, p. 437-449.

Research output: Contribution to journalArticle

@article{1766481e76504e2db41d06e4b13bb44b,
title = "Empirical Bayes two-action problem for the continuous one-parameter exponential family with errors in variables",
abstract = "In this paper we study the empirical Bayes linear-loss two-action problem for the continuous one-parameter exponential family when the observed data are contaminated (errors in variables). A new empirical Bayes testing rule is constructed, and its asymptotic optimality uniformly over a class of prior distributions is established. Uniform rates of convergence of the corresponding regret (excess risk), which depends on the type of the error distribution, are also obtained for two types of error distributions. Our results are compared with the 'pure' observed data results of the literature.",
keywords = "Asymptotically optimal, Empirical Bayes, Errors in variables, Rates of convergence",
author = "Karunamuni, {Rohana J.} and Shunpu Zhang",
year = "2003",
month = "5",
day = "1",
doi = "10.1016/S0378-3758(02)00108-8",
language = "English (US)",
volume = "113",
pages = "437--449",
journal = "Journal of Statistical Planning and Inference",
issn = "0378-3758",
publisher = "Elsevier",
number = "2",

}

TY - JOUR

T1 - Empirical Bayes two-action problem for the continuous one-parameter exponential family with errors in variables

AU - Karunamuni, Rohana J.

AU - Zhang, Shunpu

PY - 2003/5/1

Y1 - 2003/5/1

N2 - In this paper we study the empirical Bayes linear-loss two-action problem for the continuous one-parameter exponential family when the observed data are contaminated (errors in variables). A new empirical Bayes testing rule is constructed, and its asymptotic optimality uniformly over a class of prior distributions is established. Uniform rates of convergence of the corresponding regret (excess risk), which depends on the type of the error distribution, are also obtained for two types of error distributions. Our results are compared with the 'pure' observed data results of the literature.

AB - In this paper we study the empirical Bayes linear-loss two-action problem for the continuous one-parameter exponential family when the observed data are contaminated (errors in variables). A new empirical Bayes testing rule is constructed, and its asymptotic optimality uniformly over a class of prior distributions is established. Uniform rates of convergence of the corresponding regret (excess risk), which depends on the type of the error distribution, are also obtained for two types of error distributions. Our results are compared with the 'pure' observed data results of the literature.

KW - Asymptotically optimal

KW - Empirical Bayes

KW - Errors in variables

KW - Rates of convergence

UR - http://www.scopus.com/inward/record.url?scp=0037408785&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0037408785&partnerID=8YFLogxK

U2 - 10.1016/S0378-3758(02)00108-8

DO - 10.1016/S0378-3758(02)00108-8

M3 - Article

AN - SCOPUS:0037408785

VL - 113

SP - 437

EP - 449

JO - Journal of Statistical Planning and Inference

JF - Journal of Statistical Planning and Inference

SN - 0378-3758

IS - 2

ER -