Abstract
The literature on variable selection for mean regression is quite rich, both in the classical as well as in the Bayesian setting. However, if the goal is to assess the effects of the predictors at different levels of the response variable then quantile regression is useful. In this paper, we develop a Bayesian variable selection method for longitudinal response at some prefixed quantile levels of the response. We consider an Asymmetric Laplace Distribution (ALD) for the longitudinal response, and develop a simple Gibbs sampler algorithm for variable selection at each quantile level. We analyze a dataset from the health and retirement study (HRS) conducted by the University of Michigan for understanding the relationship between the physical health and the financial health of the aged individuals. We consider the out-of-pocket medical expenses as our response variable since it summarizes the physical and the financial well-being of an aged individual. Our proposed approach efficiently selects the important predictors at different prefixed quantile levels. Simulation studies are performed to assess the practical usefulness of the proposed approach. We also compare the performance of the proposed approach to some other existing methods of variable selection in quantile regression.
Similar content being viewed by others
References
Albert JH, Chib S (1993) Bayesian analysis of binary and polychotomous response data. J Am Stat Assoc 88:669–679
Al-Hamzawi R, Yu K, Benoit D (2012) Bayesian adaptive lasso quantile regression. Stat Model 12:279–297
Benoit D, Al-Hamzawi R, Yu K (2013) Bayesian lasso binary quantile regression. Comput Stat 28:2861–2873
Biswas J, Das K (2019) A Bayesian approach of analyzing semi-continuous longitudinal data with monotone missingness. Stat Model 20:148–170
Biswas J, Ghosh P, Das K (2020) A semi-parametric quantile regression approach to zero-inflated and incomplete longitudinal outcomes. Adv Stat Anal 104:261–283
Biswas J, Das K (2021) A Bayesian quantile regression approach to multivariate semi-continuous longitudinal data. Comput Stat 36:241–260
Brooks SP, Gelman A (1998) General methods for monitoring convergence of iterative simulations. J Comput Graph Stat 7:434–455
Cai B, Dunson DB (2008) Bayesian variable selection in generalized linear mixed models. Random effect and latent variable model selection. pp 63–91. Springer, New York
Das K, Ghosh P, Daniels MJ (2021) Modeling multiple time-varying related groups: a dynamic hierarchical Bayesian approach with an application to the health and retirement study. J Am Stat Assoc 116:558–568
Das K, Pareek B, Brown S, Ghosh P (2021) A semi-parametric Bayesian dynamic hurdle model with an application to the health and retirement study. Comput Stat (published online). https://doi.org/10.1007/s00180-021-01143-x
Feng Y, Chen Y, He X (2015) Bayesian quantile regression with approximate likelihood. Bernoulli 21:832–850
Fernandez C, Ley E, Steel MF (2001) Benchmark priors for Bayesian model averaging. J Econ 100:381–427
George EI, McCulloch RE (1993) Variable selection via Gibbs sampling. J Am Stat Assoc 88:881–889
George EI, McCulloch RE (1997) Approaches for Bayesian variable selection. Stat Sin 7:339–373
Geraci M, Bottai M (2007) Quantile regression for longitudinal data using the asymmetric Laplace distribution. Biostatistics 8:140–154
Jang W, Wang HJ (2015) A semiparametric Bayesian approach for joint-quantile regression with clustered data. Comput Stat Data Anal 84:99–115
Koenker R, Bassett G (1978) Regression quantiles. Econometrica 46:33–50
Koenker R (2004) Quantile regression for longitudinal data. J Multivar Anal 91:74–89
Kong Y, Li Y, Zerom D (2019) Screening and selection for quantile regression using an alternative measure of variable importance. J Multivar Anal 173:435–455
Kozumi H, Kobayashi G (2011) Gibbs sampling methods for Bayesian quantile regression. J Stat Comput Simul 81:1565–1578
Kulkarni H, Biswas J, Das K (2018) A joint quantile regression model for multiple longitudinal outcomes. Adv Stat Anal 103:453–473
Kuo L, Mallick B (1998) Variable selection for regression models. Sankhyā Indian J Stat Ser B 60:65–81
Laird NM, Ware JH (1982) Random-effects models for longitudinal data. Biometrics 38:963–974
Li Q, ** R, Lin N (2010) Bayesian regularized quantile regression. Bayesian Anal 5:533–556
Marino MF, Farcomeni A (2015) Linear quantile regression models for longitudinal experiments: an overview. Metron 73:229–247
Mukherji A, Roychoudhury S, Ghosh P, Brown S (2016) Estimating health demand for an aging population: a flexible and robust bayesian joint model. J Appl Econ 31:1140–1158
Park T, Casella G (2008) The bayesian lasso. J Am Stat Assoc 103:681–686
Reich B, Smith L (2013) Bayesian quantile regression for censored data. Biometrics 69:651–660
Sala-i-Martin X, Doppelhofer G, Miller RI (2004) Determinants of long-term growth: A Bayesian averaging of classical estimates (BACE) approach. Am Econ Rev, pp 813–835
Smith M, Kohn R (1996) Nonparametric regression using Bayesian variable selection. J Econ 75:317–344
Taddy M, Kottas A (2010) A Bayesian nonparametric approach to inference for quantile regression. J Am Stat Assoc 28:357–369
Tokdar S, Kadane J (2012) Simultaneous linear quantile regression: a semiparametric Bayesian approach. Bayesian Anal 7:51–72
Wu Y, Liu Y (2009) Variable selection in quantile regression. Stat Sin 19:801–817
Wichitaksorn N, Choy S, Gerlach R (2014) A generalized class of skew distributions and associated robust quantile regression models. Can J Stat 42:579–596
Yang Y, Tokdar S (2017) Joint estimation of quantile planes over arbitrary predictor space. J Am Stat Assoc 112:1107–1120
Yu K, Moyeed R (2001) Bayesian quantile regression. Stat Probab Lett 54:437–447
Yu K, Chen CW, Reed C, Dunson DB (2013) Bayesian variable selection in quantile regression. Stat Interface 6:261–274
Funding
No funding information is applicable for this work.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare no conflict of interest/competing interests.
Availability of data and material
The HRS dataset and the relevant information on it are all available with this manuscript.
Code availability
Relevant R code is available with this manuscript.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
About this article
Cite this article
Kedia, P., Kundu, D. & Das, K. A Bayesian variable selection approach to longitudinal quantile regression. Stat Methods Appl 32, 149–168 (2023). https://doi.org/10.1007/s10260-022-00645-2
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10260-022-00645-2