Two Stage Approach to Bayes Sequential Estimation in the Exponential Distribution

碩士 === 淡江大學 === 數學學系 === 86 === The Bayes sequential estimation problem is to seek an optimal sequential procedure Which includes an optimal stopping time and a Bayes estimator. (i.e. the posterior mean of θ for square error loss.) The...

Full description

Bibliographic Details
Main Authors: Hwang, Bang-Hwang, 黃榜煌
Other Authors: Hwang Leng-Cheng
Format: Others
Language:zh-TW
Published: 1998
Online Access:http://ndltd.ncl.edu.tw/handle/03593077172527607183
id ndltd-TW-086TKU01479017
record_format oai_dc
spelling ndltd-TW-086TKU014790172015-10-13T17:34:46Z http://ndltd.ncl.edu.tw/handle/03593077172527607183 Two Stage Approach to Bayes Sequential Estimation in the Exponential Distribution 貝氏序列下指數分佈的二階段抽樣法 Hwang, Bang-Hwang 黃榜煌 碩士 淡江大學 數學學系 86 The Bayes sequential estimation problem is to seek an optimal sequential procedure Which includes an optimal stopping time and a Bayes estimator. (i.e. the posterior mean of θ for square error loss.) The problem of determining optimal stopping rules is usually more formidable, and although such stopping rules exist under fairly general conditions, their exact determination is usually very difficult. There are many papers that discussed the optimal stopping time. For example, Chow, Robbins and Siegmund (1971) proved the existence of the optimal stopping time for a Bayes sequential estimation problem. Alvo (1977) obtained a lower bound for a Bayes risk of optimal stopping rules for one-parameter exponential family. Of course, there are many papers that discussed the asymptotic efficiency for the stopping time. That is, its Bayes risk compare to the Baye riskof optimal stopping time. In the classical non-Bayesian sequential estimation, Robbins (1959) proposed a sequential procedure and compared its Bayes risk with the Bayes of the optimal fixed sample size procedure in normal case. Replacing of fully sequential procedure, a two stage procedure is given by Ghosh and Mukhopadyay (1981). The asymptotic properties of the two stage procedure were discussed. In the Bayes sequential estimation problem the optimal fixed sample size procedure depends on prior distribution. In case the prior distribution is misspecified, unknown or irrelevant to the present θ, the optiaml fixed sample size can not be obtained. Hwang (1997) proposed a naive sequential procedure and did not depend on the prior distribution such that its Bayes risk was not greater than its Bayes risk of optimal fixed sample size procedure. Instead of the sequential procedure in Hwang (1997), a two stage procedure is proposed in the present paper. The asymptotic properties of procedure are the same as the sequential procedure proposed by Hwang (1997),a two stage procedure is proposed in the present paper. The asymptotic properties of procedure are the same as the sequential procedure proposed by Hwang (1997). Hwang Leng-Cheng 黃連成 1998 學位論文 ; thesis 28 zh-TW
collection NDLTD
language zh-TW
format Others
sources NDLTD
description 碩士 === 淡江大學 === 數學學系 === 86 === The Bayes sequential estimation problem is to seek an optimal sequential procedure Which includes an optimal stopping time and a Bayes estimator. (i.e. the posterior mean of θ for square error loss.) The problem of determining optimal stopping rules is usually more formidable, and although such stopping rules exist under fairly general conditions, their exact determination is usually very difficult. There are many papers that discussed the optimal stopping time. For example, Chow, Robbins and Siegmund (1971) proved the existence of the optimal stopping time for a Bayes sequential estimation problem. Alvo (1977) obtained a lower bound for a Bayes risk of optimal stopping rules for one-parameter exponential family. Of course, there are many papers that discussed the asymptotic efficiency for the stopping time. That is, its Bayes risk compare to the Baye riskof optimal stopping time. In the classical non-Bayesian sequential estimation, Robbins (1959) proposed a sequential procedure and compared its Bayes risk with the Bayes of the optimal fixed sample size procedure in normal case. Replacing of fully sequential procedure, a two stage procedure is given by Ghosh and Mukhopadyay (1981). The asymptotic properties of the two stage procedure were discussed. In the Bayes sequential estimation problem the optimal fixed sample size procedure depends on prior distribution. In case the prior distribution is misspecified, unknown or irrelevant to the present θ, the optiaml fixed sample size can not be obtained. Hwang (1997) proposed a naive sequential procedure and did not depend on the prior distribution such that its Bayes risk was not greater than its Bayes risk of optimal fixed sample size procedure. Instead of the sequential procedure in Hwang (1997), a two stage procedure is proposed in the present paper. The asymptotic properties of procedure are the same as the sequential procedure proposed by Hwang (1997),a two stage procedure is proposed in the present paper. The asymptotic properties of procedure are the same as the sequential procedure proposed by Hwang (1997).
author2 Hwang Leng-Cheng
author_facet Hwang Leng-Cheng
Hwang, Bang-Hwang
黃榜煌
author Hwang, Bang-Hwang
黃榜煌
spellingShingle Hwang, Bang-Hwang
黃榜煌
Two Stage Approach to Bayes Sequential Estimation in the Exponential Distribution
author_sort Hwang, Bang-Hwang
title Two Stage Approach to Bayes Sequential Estimation in the Exponential Distribution
title_short Two Stage Approach to Bayes Sequential Estimation in the Exponential Distribution
title_full Two Stage Approach to Bayes Sequential Estimation in the Exponential Distribution
title_fullStr Two Stage Approach to Bayes Sequential Estimation in the Exponential Distribution
title_full_unstemmed Two Stage Approach to Bayes Sequential Estimation in the Exponential Distribution
title_sort two stage approach to bayes sequential estimation in the exponential distribution
publishDate 1998
url http://ndltd.ncl.edu.tw/handle/03593077172527607183
work_keys_str_mv AT hwangbanghwang twostageapproachtobayessequentialestimationintheexponentialdistribution
AT huángbǎnghuáng twostageapproachtobayessequentialestimationintheexponentialdistribution
AT hwangbanghwang bèishìxùlièxiàzhǐshùfēnbùdeèrjiēduànchōuyàngfǎ
AT huángbǎnghuáng bèishìxùlièxiàzhǐshùfēnbùdeèrjiēduànchōuyàngfǎ
_version_ 1717781221107302400