Estimation of the Weibull Distribution under Unified Progressive Hybrid Censored Data

• 발행기관: 한국자료분석학회
• 저자명: Junseok Kim Kyeongjun Lee
• 간행물 정보: 『Journal of The Korean Data Analysis Society (JKDAS)』Vol.20 No.5, 2189~2199쪽, 전체 11쪽
• 주제분류: 자연과학 > 통계학
• 파일형태: PDF
• 발행일자: 2018.10.31

영문 초록

It is usually recognized that the life-times of test items may not be registered exactly. Recently, unified progressive hybrid censoring (UPHC) scheme have been introduced. In this paper, therefore, we consider the maximum likelihood estimates (MLEs) and Bayes estimates (BEs) of the parameters and reliability of a Weibull distribution (WD) under UPHC scheme. For computing the MLEs of the parameters, we use the Kundu’s algorithm. Further, the Bayes estimates for the parameters and reliability of WD based on squared error loss function (SELF) are derived. Since the BEs cannot be derived in closed form, we derive the BEs by using the Lindley approximation method. We carried out Monte Carlo (MC) simulations to compare the MLEs and BEs. It is observed that the BEs of the parameters and reliability under the SELF work quite well in this work. In addition, real data set based on UPHC scheme have been also analysed for illustrative purpose.

목차

1. Introduction

2. Unified progressive hybrid censoring

3. Maximum likelihood estimation

4. Bayes estimation

5. Illustrative example and simulation results

6. Conclusions

References

키워드

Lindley approximation maximum likelihood estimation unified progressive hybrid censoring Weibull distribution