论文信息 - Bayesian Estimation of Measurement Error Models with Longitudinal Data

Bayesian Estimation of Measurement Error Models with Longitudinal Data

In this paper, Bayesian inferences for semiparametric measurement error models (MEs) for longitudinal data are investigated. A semiparametric Bayesian approach combining the stick-breaking prior and the Gibbs sampler together with the Metropolis-Hastings algorithm is developed for simulating observations from the posterior distributions and producing the joint Bayesian estimates of unknow parameters and measurement error. We obtain Bayesian estimations of parameters and covariates subject to MEs. Two simulation studies are presented to illustrate our proposed methodologies. Keywords-Bayesian; measurement error; longitudinal data

Dewang Li | Meilan Qiu | Dewang Li | Meilan Qiu

[1] Raymond J. Carroll,et al. Bias Analysis and SIMEX Approach in Generalized Linear Mixed Measurement Error Models , 1998 .

[2] Wayne A. Fuller,et al. Measurement Error Models , 1988 .

[3] R. Carroll,et al. Locally Efficient Estimators for Semiparametric Models With Measurement Error , 2006 .

[4] Xihong Lin,et al. Estimation in Semiparametric Transition Measurement Error Models for Longitudinal Data , 2009, Biometrics.

[5] V. H. Lachos,et al. Bayesian analysis of skew-t multivariate null intercept measurement error model , 2010 .

[6] Hua Liang,et al. Partially Linear Models with Missing Response Variables and Error-prone Covariates. , 2007, Biometrika.

[7] D. Ruppert,et al. Measurement Error in Nonlinear Models , 1995 .

[8] W. Härdle,et al. Estimation in a semiparametric partially linear errors-in-variables model , 1999 .

[9] Donald Geman,et al. Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images , 1984 .