CHARACTERIZATION OF ADMISSIBLE LINEAR ESTIMATORS UNDER EXTENDED BALANCED LOSS FUNCTION

Mirezi B., KAÇIRANLAR S.

KYBERNETIKA, vol.57, no.4, pp.613-627, 2021 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 57 Issue: 4
  • Publication Date: 2021
  • Doi Number: 10.14736/kyb-2021-4-0613
  • Journal Name: KYBERNETIKA
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Linguistic Bibliography, zbMATH
  • Page Numbers: pp.613-627
  • Keywords: admissibility, extended balanced loss function, linear admissible estimator, REGRESSION COEFFICIENT
  • Çukurova University Affiliated: Yes

Abstract

In this paper, we study the admissibility of linear estimator of regression coefficient in linear model under the extended balanced loss function (EBLF). The sufficient and necessary condition for linear estimators to be admissible are obtained respectively in homogeneous and non-homogeneous classes. Furthermore, we show that admissible linear estimator under the EBLF is a convex combination of the admissible linear estimator under the sum of square residuals and quadratic loss function.