Please use this identifier to cite or link to this item:
https://ah.lib.nccu.edu.tw/handle/140.119/84953| 題名: | 處理失去部分訊息資料問題的準貝氏法 Quasi-Bayesian methods on the problem with censored data |
作者: | 劉猷銘 | 貢獻者: | 姜志銘 劉猷銘 |
關鍵詞: | 準貝式法 quasi-Bayes method |
日期: | 2001 | 上傳時間: | 15-Apr-2016 | 摘要: | 以貝式方法處理部分區分(partially-classified)失去部分訊息資料的類別抽樣(categorical sampling with censored data)大部分皆建立在“無價值性失去部分訊息的類別資料”(non-informative censored categorical data)與“誠實回答”(truthful reporting)的前提下。Jiang (1995)取消這兩個假設，提出類似Makov & Smith(1977)和Smith & Makov(1978)對混合分配(mixture distribution)所用之準貝式法(quasi-Bayes method)來得到近似解。而本文將討Jiang提出之準貝式法的收斂性，及考慮先驗分配對估計精準度的影響。 封面頁\r\n證明書\r\n致謝詞\r\n論文摘要\r\n目錄\r\n1 簡介\r\n2 準貝式法\r\n2.1 多元白努利抽樣\r\n2.2 準貝式法\r\n3 準貝式法之收斂性\r\n3.1 某些先驗分配下的結果\r\n3.2 收斂性質的模擬\r\n4 模擬先驗分配對估計精準度之影響\r\n5 結論\r\nReference\r\n附錄\r\n附錄1：收斂性質的模擬\r\n附錄2：情形一到情形七的α、β、ｕ和Ｇ\r\n附錄3：模擬先驗分配對估計精準度之影響 |
參考文獻: | Dickey,J.M., Jiang,J.M., and Kadane, J. B. (1987),\" Bayesian Methods for Censored Categorical Data,\" Journal of the American Ststistical Association, 82, 773-781.\r\nJiang,T.J.(1995) \" Quasi-Bayes Sequential for Categorical Data under Informative Censoring ,\" Technical report #1995-02, Department of Mathematical Sciences, National Chengchi University.\r\nJiang,T.J.,and Dickey,J.M. (2001), \"Bayesian Approaches to Categorical Data with Informative Censoring,\" To be published.\r\nMakov, U. E., and Smith, A. F. M. (1977), \"A Quasi-Bayes Unsupervised Learning Procedures for Priors, \" IEEE Trans.Inf.Theory,IT-23,761-764.\r\nSmith, A. F. M., and Makov, U. E. (1978), \" A Quasi-Bayes Sequential Procedure for Mixture, \" Journal of the Royal Statistical Society,Ser. B. 40, 106-112\r\nVenter, J. H. (1966) \"On Dvroetzky Stochastic Approximation, \" Ann. Math. Statist, 37, 1534-1544. | 描述: | 碩士 國立政治大學 應用數學系 |
資料來源: | http://thesis.lib.nccu.edu.tw/record/#A2002001144 | 資料類型: | thesis |
| Appears in Collections: | 學位論文 |
Files in This Item:
| File | Size | Format | |
|---|---|---|---|
| index.html | 115 B | HTML2 | View/Open |
Google ScholarTM
Check
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.