Published
Global Variable Selection for Quantile Regression
Selección global de variables para regresión cuantílica
DOI:
https://doi.org/10.15446/rce.v48n3.122651Keywords:
Chebyshev polynomials, Group adaLASSO. (en)Polinomios de Chebyshev, Group adaLASSO. (es)
Downloads
Quantile regression provides a parsimonious model for the conditional quantile function of the response variable Y given the vector of covariates X, and describes the whole conditional distribution of the response, yielding estimators that are more robust to the presence of outliers. Quantile regression models specify, for each quantile level τ , the functional form for the conditional τ -th quantile of the response, which brings complexity to perform variable selection using regularization techniques, such as LASSO or adaptive LASSO (adaLASSO), as one might obtain a different set of selected variables for each quantile level. In this work, we propose a method for global variable selection and coefficient estimation in the linear quantile regression framework, imposing few restrictions on the functional form of β(·), and applying group adaLASSO penalization for variable selection. We set up a Monte Carlo study comparing six different proposed estimators based on LASSO, adaLASSO and group LASSO in six scenarios that diversify sample and quantile levels grid sizes. The findings demonstrate that the selection of the tuning parameter λ for penalization is critical for model selection and coefficient estimation. It was observed that the methods using traditional LASSO are more prone to include the true model as compared to adaLASSO, but renouncing model shrinkage and not removing irrelevant covariates, while the grouped approaches are more effective in zeroing coefficients that are less relevant.
La regresión cuantílica proporciona un modelo parsimonioso para la función de cuantiles condicionales de la variable de respuesta Y dado el vector de covariables X, y describe toda la distribución condicional de la respuesta, produciendo estimadores más robustos ante la presencia de valores atípicos. Los modelos de regresión cuantílica especifican, para cada nivel de cuantil τ , la forma funcional del τ -ésimo cuantil condicional de la respuesta, lo que introduce complejidad a la hora de realizar selección de variables mediante técnicas de regularización, como LASSO o LASSO adaptativo (adaLASSO), ya que podríamos obtener un conjunto diferente de variables seleccionadas para cada nivel de cuantil. En este trabajo proponemos un método de selección global de variables y estimación de coeficientes en el marco de la regresión cuantílica lineal, imponiendo pocas restricciones sobre la forma funcional de β(·) y aplicando una penalización adaLASSO por grupos para la selección de variables. Realizamos un estudio de Monte Carlo comparando seis estimadores propuestos basados en LASSO, adaLASSO y LASSO grupal en seis escenarios que diversifican los tamaños de muestra y de la rejilla de niveles de cuantil. Los resultados demuestran que la selección del parámetro de ajuste λ para la penalización es crítica para la selección del modelo y la estimación de coeficientes. Se observó que los métodos que utilizan LASSO tradicional son más propensos a incluir el modelo verdadero en comparación con adaLASSO, pero a costa de renunciar a la contracción del modelo y de no eliminar covariables irrelevantes, mientras que los enfoques agrupados son más eficaces para anular (llevar a cero) los coeficientes menos relevantes.
References
Akaike, H. (1973), Information Theory and an Extension of the Maximum Likelihood Principle, Springer New York, New York, NY, pp. 199-213.
Azimli, A. (2020), ‘The impact of covid-19 on the degree of dependence and structure of risk-return relationship: A quantile regression approach’, Finance Research Letters 36, 101648.
Bach, F., Jenatton, R., Mairal, J. & Obozinski, G. (2012), ‘Optimization with sparsity-inducing penalties’, Foundations and Trendső in Machine Learning 4(1), 1-106.
Belloni, A. & Chernozhukov, V. (2011), ‘l1-penalized quantile regression in highdimensional sparse models’, The Annals of Statistics 39(1), 82-130.
Bishop, C. M. (2006), Pattern Recognition and Machine Learning (Information Science and Statistics), Springer-Verlag, Berlin, Heidelberg, p. 217.
Bonaccorsi, G., Pierri, F., Cinelli, M., Flori, A., Galeazzi, A., Porcelli, F., Schmidt, A. L., Valensise, C. M., Scala, A., Quattrociocchi, W. & Pammolli, F. (2020), ‘Economic and social consequences of human mobility restrictions under covid19’, Proceedings of the National Academy of Sciences 117(27), 15530-15535.
Buchinsky, M. (1998), ‘Recent advances in quantile regression models: A practical guideline for empirical research’, Journal of Human Resources 33(1), 88-126.
Cattaneo, M. D., Chandak, R., Jansson, M. & Ma, X. (2022), Boundary adaptive local polynomial conditional density estimators, arXiv preprint 2204.10359, arXiv.
Das, P. & Ghosal, S. (2018), ‘Bayesian non-parametric simultaneous quantile regression for complete and grid data’, Computational Statistics & Data Analysis 127, 172–186.
Davino, C., Furno, M. & Vistocco, D. (2014), Quantile Regression: Theory and Applications, Wiley.
Fan, J., Yao, Q. & Tong, H. (1996), ‘Estimation of conditional densities and sensitivity measures in nonlinear dynamical systems’, Biometrika 83(1), 189-206.
Frumento, P. & Bottai, M. (2016), ‘Parametric modeling of quantile regression coefficient functions’, Biometrics 72(1), 74-84.
Frumento, P., Bottai, M. & Fernández-Val, I. (2021), ‘Parametric modeling of quantile regression coefficient functions with longitudinal data’, Journal of the American Statistical Association 116(534), 783-797.
Frumento, P. & Salvati, N. (2021), ‘Parametric modeling of quantile regression coefficient functions with count data’, Statistical Methods & Applications 30(4), 1237-1258.
Fu, A., Narasimhan, B. & Boyd, S. (2020), ‘CVXR: An R package for disciplined convex optimization’, Journal of Statistical Software 94(14), 1-34.
Gneiting, T. (2011), ‘Making and evaluating point forecasts’, Journal of the American Statistical Association 106(494), 746-762.
Hashem, H., Vinciotti, V., Alhamzawi, R. & Yu, K. (2016), ‘Quantile regression with group lasso for classification’, Advances in Data Analysis and Classification 10(3), 375-390.
Hunter, D. R. & Lange, K. (2000), ‘Quantile regression via an MM algorithm’, Journal of Computational and Graphical Statistics 9(1), 60-77.
Koenker, R. (2000), ‘Galton, Edgeworth, Frisch, and prospects for quantile regression in econometrics’, Journal of Econometrics 95(2), 347-374.
Koenker, R. (2004), ‘Quantile regression for longitudinal data’, Journal of Multivariate Analysis 91(1), 74–89. Cunha Medeiros, Marcelo and Mendes, Eduardo Fonseca, L_1-Regularization of High-Dimensional Time-Series Models with Flexible Innovations (July 3, 2015). Available at SSRN: https://ssrn.com/abstract=2626507 or http://dx.doi.org/10.2139/ssrn.2626507
Koenker, R. (2005), Quantile regression, Econometric Society Monographs, Cambridge University Press.
Koenker, R. (2021), quantreg: Quantile Regression. R package version 5.86. https://CRAN.R-project.org/package=quantreg
Koenker, R. & Bassett, G. (1978), ‘Regression Quantiles’, Econometrica 46(1), 33-50.
Koenker, R. & Hallock, K. F. (2001), ‘Quantile regression’, Journal of Economic Perspectives 15(4), 143-156.
Konzen, E. & Ziegelmann, F. A. (2016), ‘LASSO-Type Penalties for Covariate Selection and Forecasting in Time Series’, Journal of Forecasting 35(7), 592-612.
Li, Q., Lin, N. & Xi, R. (2010), ‘Bayesian regularized quantile regression’, Bayesian Analysis 5(3), 533-556.
Li, Y. & Zhu, J. (2008), ‘L1-norm quantile regression’, Journal of Computational and Graphical Statistics 17(1), 163-185.
Lu, H., Nie, P. & Qian, L. (2020), Do Quarantine Experiences and Attitudes Towards COVID-19 Affect the Distribution of Psychological Outcomes in China? A Quantile Regression Analysis, GLO Discussion Paper 512, Global Labor Organization, Essen.
Mairal, J., Bach, F. & Ponce, J. (2014), ‘Sparse modeling for image and vision processing’, Foundations and Trendső in Computer Graphics and Vision 8(2- 3), 85-283.
Man, R., Pan, X., Tan, K. M. & Zhou, W.-X. (2022), ‘A unified algorithm for penalized convolution smoothed quantile regression’.
Medeiros, M. C. & Mendes, E. F. (2015), l1-Regularization of High-Dimensional Time-Series Models with Flexible Innovations, Cunha Medeiros, Marcelo and Mendes, Eduardo Fonseca, L_1-Regularization of High-Dimensional Time-Series Models with Flexible Innovations (July 3, 2015). Available at SSRN: https://ssrn.com/abstract=2626507 or http://dx.doi.org/10.2139/ssrn.2626507
Park, S. & He, X. (2017), ‘Hypothesis testing for regional quantiles’, Journal of Statistical Planning and Inference 191, 13-24.
Quarteroni, A. & Valli, A. (1994), Numerical Approximation of Partial Differential Equations, Lecture Notes in Mathematics, Springer-Verlag.
Ruas, M., Street, A. & Fernandes, C. (2022), ‘A multi-quantile regression time series model with interquantile Lipschitz regularization for wind power probabilistic forecasting’, Electric Power Systems Research 209, 107973.
Schwarz, G. (1978), ‘Estimating the dimension of a model’, The Annals of Statistics 6(2), 461-464.
Sottile, G. & Frumento, P. (2022), ‘Robust estimation and regression with parametric quantile functions’, Computational Statistics and Data Analysis 171, 107471.
Sottile, G., Frumento, P., Chiodi, M. & Bottai, M. (2020), ‘A penalized approach to covariate selection through quantile regression coefficient models’, Statistical Modelling 20(4), 369-385.
Spady, R. & Stouli, S. (2020), ‘Gaussian transforms modeling and the estimation of distributional regression functions’.
Tibshirani, R. (1996), ‘Regression shrinkage and selection via the lasso’, Journal of the Royal Statistical Society. Series B (Methodological) 58(1), 267-288.
Wang, H. & Leng, C. (2008), ‘A note on adaptive group lasso’, Computational Statistics & Data Analysis 52(12), 5277-5286.
Wu, Y. & Liu, Y. (2009), ‘Variable selection in quantile regression’, Statistica Sinica 19(2), 801-817.
Yoshida, T. (2021), ‘Quantile function regression and variable selection for sparse models’, Canadian Journal of Statistics 49.
Yuan, M. & Lin, Y. (2006), ‘Model selection and estimation in regression with grouped variables’, Journal of the Royal Statistical Society. Series B: Statistical Methodology 68(1), 49-67.
Zheng, Q., Peng, L. & He, X. (2015), ‘Globally adaptive quantile regression with ultra-high dimensional data’, The Annals of Statistics 43(5), 2225-2258.
Zou, H. (2006), ‘The Adaptive Lasso and Its Oracle Properties’, Journal of the American Statistical Association 101(476), 1418–1429.
How to Cite
APA
ACM
ACS
ABNT
Chicago
Harvard
IEEE
MLA
Turabian
Vancouver
Download Citation
License

This work is licensed under a Creative Commons Attribution 4.0 International License.
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).






