東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Bayesian analyses of Markov chains: ...

Sung, Minje.

Linked to FindBook

Google Book

Amazon

博客來

Bayesian analyses of Markov chains: Applications to longitudinal data from psychiatric treatment programs.

Record Type:	Language materials, printed : Monograph/item
Title/Author:	Bayesian analyses of Markov chains: Applications to longitudinal data from psychiatric treatment programs./
Author:	Sung, Minje.
Description:	156 p.
Notes:	Director: Refik Soyer.
Contained By:	Dissertation Abstracts International62-04B.
Subject:	Health Sciences, Health Care Management. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3012326
ISBN:	0493218017

Bayesian analyses of Markov chains: Applications to longitudinal data from psychiatric treatment programs.
Sung, Minje.

Bayesian analyses of Markov chains: Applications to longitudinal data from psychiatric treatment programs. - 156 p.

Director: Refik Soyer.

Thesis (Ph.D.)--The George Washington University, 2001.

The Bayesian models and inference procedures are implemented to the real data from the psychiatric treatment study.

ISBN: 0493218017Subjects--Topical Terms:

1017922
Health Sciences, Health Care Management.

Bayesian analyses of Markov chains: Applications to longitudinal data from psychiatric treatment programs.
LDR:03191nam 2200337 a 45 001 929189
005 20110427
008 110427s2001 eng d
020 $a 0493218017
035 $a (UnM)AAI3012326
035 $a AAI3012326
040 $a UnM $c UnM
100 1 $a Sung, Minje. $3 1252674
245 1 0 $a Bayesian analyses of Markov chains: Applications to longitudinal data from psychiatric treatment programs.
300 $a 156 p.
500 $a Director: Refik Soyer.
500 $a Source: Dissertation Abstracts International, Volume: 62-04, Section: B, page: 1794.
502 $a Thesis (Ph.D.)--The George Washington University, 2001.
520 $a The Bayesian models and inference procedures are implemented to the real data from the psychiatric treatment study.
520 $a The objectives of this dissertation research are to develop Bayesian methods for modeling and analyses of Markov chains, and to develop inference procedures to be able to address issues encountered in the analyses of data from psychiatric treatment programs. Two classes of models are considered for describing the transition probabilities of a Markov chain. The first class of models is based on the Dirichlet type Bayesian models and the second class is based on the logistic regression type models. A reparameterization of the Dirichlet distribution is introduced to incorporate covariate effects into the analysis. This new modeling framework, referred to as the generalized Dirichlet model, unifies the existing models in the literature that are used for describing transition probabilities. This unification is achieved in a hierarchical Bayes setup where choice of different forms of prior distributions results in different models for transition probabilities. Extensions of the model are introduced to incorporate random effects, stochastic exits, and ordinal data into Markov chains.
520 $a Under the generalized Dirichlet model setup, a formal treatment of nonhomogeneous Markov chains is presented to describe the time evolution of transition probabilities. In so doing, two modeling strategies are introduced to describe the time dependence. The first strategy is based on the concept of exchangeability whereas the second one is based on a first order Markov property. These strategies are incorporated into Markov chain models by either using logit transforms of the mean transition probabilities in the generalized Dirichlet model or using logit transforms of the transition probabilities in the logistic regression representation.
520 $a Bayesian inference procedures are developed for the proposed models for homogeneous and nonhomogeneous Markov chains. These are considered under the generalized Dirichlet as well as the logistic regression setups. The complicated nature of these models requires the use of Markov chain Monte Carlo methods such as the Gibbs sampler.
590 $a School code: 0075.
650 4 $a Health Sciences, Health Care Management. $3 1017922
650 4 $a Health Sciences, Mental Health. $3 1017693
650 4 $a Operations Research. $3 626629
650 4 $a Statistics. $3 517247
690 $a 0347
690 $a 0463
690 $a 0769
690 $a 0796
710 2 0 $a The George Washington University. $3 1017405
773 0 $t Dissertation Abstracts International $g 62-04B.
790 $a 0075
790 1 0 $a Soyer, Refik, $e advisor
791 $a Ph.D.
792 $a 2001
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3012326