Volume no :14, Issue no: 1, September (2015)

Journal of Statistics: Advances in Theory and Applications

BAYESIAN AND NON-BAYESIAN ESTIMATION FROM WEIBULL-UNIFORM DISTRIBUTION BASED ON UPPER RECORD VALUES

Author's: A. Amin Essam and A. Abdel-Aziz Alaa
Pages: [107] - [123]
Received Date: September 16, 2015
Submitted by:
DOI: http://dx.doi.org/10.18642/jsata_7100121541

Abstract

This paper develops a Bayesian and non-Bayesian analysis in the context of upper record values from the Weibull-Uniform (WU) distribution. The maximum likelihood estimates and Bayes estimates based on upper record values are derived for the unknown parameters. The Bayes estimates are obtained with respect to both squared error loss and zero one loss functions. To compare the different estimates, the relative absolute error and mean squared error are calculated through simulation study.

Keywords

Bayesian and non-Bayesian estimation, maximum likelihood estimation, squared error loss, zero one loss, Weibull-uniform distribution, upper record values, relative absolute error.