Bayesian Shrinkage Estimation Based On Rayleigh type-II Censored Data

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 2

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/308710016

Bayesian shrinkage estimation based on Rayleigh type-II


censored data

Article in Communication in Statistics- Theory and Methods · August 2017


DOI: 10.1080/03610926.2016.1222440

CITATIONS READS

5 49

1 author:

Mehran Naghizadeh Qomi


University of Mazandaran
33 PUBLICATIONS 75 CITATIONS

SEE PROFILE

All content following this page was uploaded by Mehran Naghizadeh Qomi on 14 December 2017.

The user has requested enhancement of the downloaded file.


COMMUNICATIONS IN STATISTICS—THEORY AND METHODS
, VOL. , NO. , –
https://doi.org/./..

Bayesian shrinkage estimation based on Rayleigh type-II


censored data
M. Naghizadeh Qomi
Department of Statistics, University of Mazandaran, Babolsar, Iran

ABSTRACT ARTICLE HISTORY


In this paper, we construct a Bayes shrinkage estimator for the Rayleigh Received  May 
scale parameter based on censored data under the squared log error loss Accepted  July 
Downloaded by [mehran naghizadeh qomi] at 01:57 26 August 2017

function. Risk-unbiased estimator is derived and its risk is computed. A


KEYWORDS
Bayes shrinkage estimator is obtained when a prior point guess value is Bayes shrinkage estimator;
available for the scale parameter. Risk-bias of the Bayes shrinkage esti- censored data; Rayleigh
mator is considered. A comparison between the proposed Bayes shrink- distribution; squared log
age estimator and the risk-unbiased estimator is provided using calcula- error loss function.
tion of the relative efficiency. A numerical example is presented for illus-
trative and comparative purposes.

1. Introduction
Let X has a one parameter Rayleigh distribution with probability density function (p.d.f.)
 
x x2
f (x; σ ) = exp − , x>0 (1)
σ 2σ
and the corresponding cumulative density function (c.d.f.) as
 
x2
F (x; σ ) = 1 − exp − , x>0 (2)

The failure rate of the Rayleigh distribution is
f (t; σ ) t
r(t; σ ) = = (3)
1 − F (t; σ ) σ
which is linear and increasing failure rate and makes it an appropriate distribution for mod-
eling the lifetime distribution of components which age rapidly with time. Several types of
electro-vacuum devices have this feature; see Polovko (1968).
In many situations, we have an initial value referred as guess value regarding the parameter
from past investigation or any other sources whatsoever, which is considered as non sample
information or uncertain prior information. Therefore, we can combine sample information
and non sample information in forming a new estimation approach in the hope that it will
perform better than the usual estimator such as maximum-likelihood estimator (MLE) or
risk-unbiased estimator. Thompson (1968) proposed new estimators by moving the unbiased

CONTACT M. Naghizadeh Qomi m.naghizadeh@umz.ac.ir Department of Statistics, University of Mazandaran,


Babolsar, Iran.
©  Taylor & Francis Group, LLC

View publication stats

You might also like