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

Parametric latent class joint model ...

Han, Jun.

Linked to FindBook

Google Book

Amazon

博客來

Parametric latent class joint model for longitudinal markers and recurrent events.

Record Type:	Electronic resources : Monograph/item
Title/Author:	Parametric latent class joint model for longitudinal markers and recurrent events./
Author:	Han, Jun.
Description:	95 p.
Notes:	Source: Dissertation Abstracts International, Volume: 66-04, Section: B, page: 2142.
Contained By:	Dissertation Abstracts International66-04B.
Subject:	Statistics. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3173161
ISBN:	0542102196

Parametric latent class joint model for longitudinal markers and recurrent events.
Han, Jun.

Parametric latent class joint model for longitudinal markers and recurrent events. - 95 p.

Source: Dissertation Abstracts International, Volume: 66-04, Section: B, page: 2142.

Thesis (Ph.D.)--University of South Carolina, 2005.

A joint model for longitudinal markers and recurrent events is proposed. This general model accommodates internal (marker) covariates, external covariates, effects of accumulating event occurrences, and effects caused by intervention after each event occurrence. A latent class mixture-of-experts model is employed to handle an underlying heterogeneous population and associated covariates. A likelihood approach is adopted to obtain parameter estimates. A BIC approximation of integrated classification likelihood criterion is used to select the number of latent classes. A joint distance of longitudinal markers and recurrent events is proposed for classification purposes. Methods of cluster analysis based on the proposed distance are compared. The EM algorithm is used to estimate the parameters in the joint model, and a bootstrap approach is adopted to construct confidence intervals for the parameters. Explicit formulas are given to compute the initial values of parameters for the EM algorithm. Prediction of class membership and event occurrence are considered. Computational issues for obtaining estimates are addressed. A simulation study is performed to examine the properties of the estimators numerically, and theoretical properties of the estimators such as existence, consistency, uniqueness, and asymptotic normality are obtained analytically.

ISBN: 0542102196Subjects--Topical Terms:

517247
Statistics.

Parametric latent class joint model for longitudinal markers and recurrent events.
LDR:02301nmm 2200277 4500 001 1817523
005 20060814143910.5
008 130610s2005 eng d
020 $a 0542102196
035 $a (UnM)AAI3173161
035 $a AAI3173161
040 $a UnM $c UnM
100 1 $a Han, Jun. $3 1294702
245 1 0 $a Parametric latent class joint model for longitudinal markers and recurrent events.
300 $a 95 p.
500 $a Source: Dissertation Abstracts International, Volume: 66-04, Section: B, page: 2142.
500 $a Directors: Edsel A. Pena; Elizabeth H. Slate.
502 $a Thesis (Ph.D.)--University of South Carolina, 2005.
520 $a A joint model for longitudinal markers and recurrent events is proposed. This general model accommodates internal (marker) covariates, external covariates, effects of accumulating event occurrences, and effects caused by intervention after each event occurrence. A latent class mixture-of-experts model is employed to handle an underlying heterogeneous population and associated covariates. A likelihood approach is adopted to obtain parameter estimates. A BIC approximation of integrated classification likelihood criterion is used to select the number of latent classes. A joint distance of longitudinal markers and recurrent events is proposed for classification purposes. Methods of cluster analysis based on the proposed distance are compared. The EM algorithm is used to estimate the parameters in the joint model, and a bootstrap approach is adopted to construct confidence intervals for the parameters. Explicit formulas are given to compute the initial values of parameters for the EM algorithm. Prediction of class membership and event occurrence are considered. Computational issues for obtaining estimates are addressed. A simulation study is performed to examine the properties of the estimators numerically, and theoretical properties of the estimators such as existence, consistency, uniqueness, and asymptotic normality are obtained analytically.
590 $a School code: 0202.
650 4 $a Statistics. $3 517247
690 $a 0463
710 2 0 $a University of South Carolina. $3 1017477
773 0 $t Dissertation Abstracts International $g 66-04B.
790 1 0 $a Pena, Edsel A., $e advisor
790 1 0 $a Slate, Elizabeth H., $e advisor
790 $a 0202
791 $a Ph.D.
792 $a 2005
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3173161