A Study of Estimation for the Three-Parameter Weibull Distribution Based on Doubly Type-II Censored Data Using a Least Squares Method

Author

Nagatsuka, Hideki

Author_Institution

Fac. of Syst. Design, Tokyo Metropolitan Univ., Hino

fYear

2008

fDate

14-17 July 2008

Firstpage

158

Lastpage

165

Abstract

A least squares procedure for parameter estimation of the three-parameter Weibull distribution based on the doubly type-2 censored samples are proposed. The estimate of the shape parameter can be obtained by minimizing a function of only one parameter and the estimates of the location and scale parameters can be obtained by two closed-formed formulas, by using the proposed method. The proposed method has advantages compared to the conventional estimators: (1) the bias of the estimators can be smaller for small samples, (2) the estimates can be obtained with higher probability, (3) the estimates are much easier to obtain computationally.

Keywords

Weibull distribution; least squares approximations; parameter estimation; least squares method; parameter estimation; three-parameter Weibull distribution; Distributed computing; Equations; Least squares approximation; Least squares methods; Maximum likelihood estimation; Moment methods; Parameter estimation; Shape; Virtual manufacturing; Weibull distribution; Censored data; Least squares method; Maximum likelihood estimation; Modefied maximum likelihood estimation; Weibull distribution;

fLanguage

English

Publisher

ieee

Conference_Titel

Secure System Integration and Reliability Improvement, 2008. SSIRI '08. Second International Conference on

Conference_Location

Yokohama

Print_ISBN

978-0-7695-3266-0

Electronic_ISBN

978-0-7695-3266-0

Type

conf

DOI

10.1109/SSIRI.2008.45

Filename

4579808