×

Flexible Bayesian quantile regression for independent and clustered data. (English) Zbl 1437.62589

Summary: Quantile regression has emerged as a useful supplement to ordinary mean regression. Traditional frequentist quantile regression makes very minimal assumptions on the form of the error distribution and thus is able to accommodate nonnormal errors, which are common in many applications. However, inference for these models is challenging, particularly for clustered or censored data. A Bayesian approach enables exact inference and is well suited to incorporate clustered, missing, or censored data. In this paper, we propose a flexible Bayesian quantile regression model. We assume that the error distribution is an infinite mixture of Gaussian densities subject to a stochastic constraint that enables inference on the quantile of interest. This method outperforms the traditional frequentist method under a wide array of simulated data models. We extend the proposed approach to analyze clustered data. Here, we differentiate between and develop conditional and marginal models for clustered data. We apply our methods to analyze a multipatient apnea duration data set.

MSC:

62P10 Applications of statistics to biology and medical sciences; meta analysis

Software:

R
PDFBibTeX XMLCite
Full Text: DOI

References:

[1] A Bayesian analysis of some nonparametric problems, 1, 209-230 (1973) · Zbl 0255.62037 · doi:10.1214/aos/1176342360
[2] Prior distribution on spaces of probability measures, 2, 615-629 (1974) · Zbl 0286.62008 · doi:10.1214/aos/1176342752
[3] Bayesian nonparametric spatial modeling with Dirichlet process mixing, 100, 1021-1035 (2005) · Zbl 1117.62342 · doi:10.1198/016214504000002078
[4] Prior distributions for variance parameters in hierarchical models (comment on article by Browne and Draper), 1, 515-534 (2006) · Zbl 1331.62139 · doi:10.1214/06-BA117A
[5] Quantile regression for longitudinal data using the asymmetric Laplace distribution, 8, 140-154 (2007) · Zbl 1170.62380 · doi:10.1093/biostatistics/kxj039
[6] Order-based dependent Dirichlet processes, 101, 179-194 (2006) · Zbl 1118.62360 · doi:10.1198/016214505000000727
[7] Modeling regression error with a mixture of Polya trees, 97, 1020-1033 (2002) · Zbl 1048.62101 · doi:10.1198/016214502388618843
[8] On choosing the centering distribution in Dirichlet process mixture models, 72, 153-162 (2005) · Zbl 1123.62025 · doi:10.1016/j.spl.2004.12.008
[9] Quantile curves without crossing, 51, 186-192 (1997)
[10] Topics in non-parametric Bayesian statistics, 455-487 (2003)
[11] Nonparametric quantile inference using Dirichlet processes, 463-492 (2007)
[12] Nonparametric estimation of convex models via mixtures, 31, 174-200 (2003) · Zbl 1018.62023 · doi:10.1214/aos/1046294461
[13] Gibbs sampling methods for stick-breaking priors, 96, 161-173 (2001) · Zbl 1014.62006 · doi:10.1198/016214501750332758
[14] Quasi-likelihood for median regression models, 91, 251-257 (1996) · Zbl 0871.62060 · doi:10.1080/01621459.1996.10476683
[15] Practical confidence intervals for regression quantiles, 14, 41-55 (2005) · doi:10.1198/106186005X27563
[16] Quantile regression for longitudinal data, 91, 74-89 (2004) · Zbl 1051.62059 · doi:10.1016/j.jmva.2004.05.006
[17] (2005)
[18] Bayesian semiparametric median regression modeling, 96, 1458-1468 (2001) · Zbl 1051.62038 · doi:10.1198/016214501753382363
[19] Bayesian nonparametric modeling in quantile regression, 36, 297-319 (2009) · Zbl 1190.62053 · doi:10.1111/sjos.2009.36.issue-2
[20] Quantile regression methods for longitudinal data with drop-outs: application to CD4 cell counts of patients infected with the human immunodeficiency virus, 46, 463-476 (1997) · Zbl 0908.62114 · doi:10.1111/rssc.1997.46.issue-4
[21] Retrospective MCMC for Dirichlet process hierarchical models, 95, 169-186 (2008) · Zbl 1437.62576 · doi:10.1093/biomet/asm086
[22] Bolus location associated with videofluoroscopic and respirodeglutometric events, 48, 21-33 (2005) · doi:10.1044/1092-4388(2005/003)
[23] R: A Language and Environment for Statistical Computing (2006)
[24] A multivariate semiparametric Bayesian spatial modeling framework for hurricane surface wind fields, 1, 249-264 (2007) · Zbl 1129.62114 · doi:10.1214/07-AOAS108
[25] Bayesian growth curves using normal mixtures with nonparametric weights, 12, 208-331 (2003) · doi:10.1198/1061860031725
[26] A nonparametric model-based approach to inference for quantile regression. Technical Report ams2007-21 (2007)
[27] A Bayesian semiparametric accelerated failure time model, 55, 477-483 (1999) · Zbl 1059.62531 · doi:10.1111/j.0006-341X.1999.00477.x
[28] Inference for censored quantile regression models in longitudinal studies, 37, 756-781 (2009) · Zbl 1162.62035 · doi:10.1214/07-AOS564
[29] Detecting differential expressions in GeneChip microarray studies: a quantile approach, 102, 104-112 (2007) · Zbl 1284.62439 · doi:10.1198/016214506000001220
[30] Quantile regression models with multivariate failure time data, 61, 151-161 (2005) · Zbl 1077.62086 · doi:10.1111/biom.2005.61.issue-1
[31] Bayesian quantile regression, 54, 437-447 (2001) · Zbl 0983.62017 · doi:10.1016/S0167-7152(01)00124-9
[32] Three-step estimation in linear mixed models with skew-t distributions, 138, 1542-1555 (2008) · Zbl 1131.62016 · doi:10.1016/j.jspi.2007.04.033
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.