國立高雄大學圖資館 |

Language: English

Back

離散型缺失資料的新插補策略 = A New Strategy for I...

國立高雄大學統計學研究所

離散型缺失資料的新插補策略 = A New Strategy for Imputing

Record Type:	Language materials, printed : monographic
Paralel Title:	A New Strategy for Imputing
Author:	徐晨軒,
Secondary Intellectual Responsibility:	國立高雄大學
Place of Publication:	[高雄市]
Published:	撰者;
Year of Publication:	2015[民104]
Description:	52面圖，表 : 30公分;
Subject:	廣義吉氏取樣
Subject:	generalized Gibbs sampling
Online resource:	http://handle.ncl.edu.tw/11296/ndltd/70871924863579650319
Notes:	104年10月31日公開
Notes:	參考書目：面48-48
Summary:	過去,我們使用完全條件分佈(full conditional distribution)模型進行吉氏取樣,其關鍵的理論是取樣過程所對應的馬可夫鏈之極限分佈為目標聯合分佈。在本研究中,我們將證明:在某些條件下,吉氏取樣可在具有不完全條件分佈的模型中進行,將此稱為廣義吉氏取樣。而且以模擬的方式在「近乎相容的(nearly compatible)不完全條件分佈」模型下,廣義吉氏取樣仍可用來估計目標聯合分佈。我們將此應用於有限離散型缺失資料的插補策略。 In the past, we use full conditional distributions to implement Gibbs sampling. This is due to the stationary distribution of the corresponding Markov chain is the target joint distribution. In this article, we will show that, under some conditions, the Gibbs sampler can be implemented by using non-full conditional distributions. We call this the generalized Gibbs sampling. In addition, based on the result of simulation, the generalized Gibbs sampling is practicable for full or non-full conditional distributions which are nearly compatible. Finally, we provide a new imputation strategy for discrete missing data.

離散型缺失資料的新插補策略 = A New Strategy for Imputing
徐, 晨軒

離散型缺失資料的新插補策略 = A New Strategy for Imputing / 徐晨軒撰 - [高雄市] : 撰者, 2015[民104]. - 52面 ; 圖，表 ; 30公分.
104年10月31日公開參考書目：面48-48.
廣義吉氏取樣generalized Gibbs sampling

離散型缺失資料的新插補策略 = A New Strategy for Imputing
LDR:02248nam0a2200289 450 001 458366
005 20170214101154.0
009 458366
010 0 $b 精裝
010 0 $b 平裝
100 $a 20170214d2015 k y0chiy50 e
101 1 $a chi $d chi $d eng
102 $a tw
105 $a ak am 000yy
200 1 $a 離散型缺失資料的新插補策略 $d A New Strategy for Imputing $z eng $f 徐晨軒撰
210 $a [高雄市] $c 撰者 $d 2015[民104]
215 0 $a 52面 $c 圖，表 $d 30公分
300 $a 104年10月31日公開
300 $a 參考書目：面48-48
314 $a 指導教授：郭錕霖博士
328 $a 碩士論文--國立高雄大學統計學研究所
330 $a 過去,我們使用完全條件分佈(full conditional distribution)模型進行吉氏取樣,其關鍵的理論是取樣過程所對應的馬可夫鏈之極限分佈為目標聯合分佈。在本研究中,我們將證明:在某些條件下,吉氏取樣可在具有不完全條件分佈的模型中進行,將此稱為廣義吉氏取樣。而且以模擬的方式在「近乎相容的(nearly compatible)不完全條件分佈」模型下,廣義吉氏取樣仍可用來估計目標聯合分佈。我們將此應用於有限離散型缺失資料的插補策略。 In the past, we use full conditional distributions to implement Gibbs sampling. This is due to the stationary distribution of the corresponding Markov chain is the target joint distribution. In this article, we will show that, under some conditions, the Gibbs sampler can be implemented by using non-full conditional distributions. We call this the generalized Gibbs sampling. In addition, based on the result of simulation, the generalized Gibbs sampling is practicable for full or non-full conditional distributions which are nearly compatible. Finally, we provide a new imputation strategy for discrete missing data.
510 1 $a A New Strategy for Imputing $z eng
610 0 $a 廣義吉氏取樣 $a 不完全條件分佈 $a 插補法 $a 缺失資料
610 1 $a generalized Gibbs sampling $a non-full conditional distribution $a imputation $a missing data
681 $a 008M/0019 $b 343201 2865 $v 2007年版
700 1 $a 徐 $b 晨軒 $4 撰 $3 709610
712 0 2 $a 國立高雄大學 $b 統計學研究所 $3 166081
801 0 $a tw $b NUK $c 20151027 $g CCR
856 7 $z 電子資源 $2 http $u http://handle.ncl.edu.tw/11296/ndltd/70871924863579650319