Abstract

Linear quantile regression model assumes quantiles of a response at

certain levels are linearly related with covariates. If the model is assumed for

one single quantile level, the semiparametric efficient estimation involves estimation of the conditional density of an error given covariates, which could be

prohibitively difficult because of the curse of dimensionality.

However, if the

model is assumed for all quantile levels, estimation of conditional density becomes estimation of the derivative of regression coefficient functions, which is

naturally available from initial estimators such as the Koenker-Bassett estimator. This paper derives the semiparametric efficient scores and the corresponding

efficiency bounds for the regression coefficients. Although there is no closed form

expression of the estimator or estimating function, we propose a computationally

feasible procedure leading to semiparametrically efficient estimation. Simulation

studies show that the proposed method could lead to substantial efficiency gain

over the standard methods.

Information

Preprint No.SS-2024-0378
Manuscript IDSS-2024-0378
Complete AuthorsZhanfeng Wang, Kani Chen, Yuanyuan Lin, Zhiliang Ying
Corresponding AuthorsKani Chen
Emailsmakchen@ust.hk

References

  1. [1] Bickel, P. J., Klaassen, C. A., Ritov, Y. and Wellner, J. A.
  2. (1993). Efficient and Adaptive Estimation for Semiparametric Models. New York: Springer-Verlag.
  3. [2] Bondell, H. D., Reich, B. J. and Wang, H. (2010). Noncrossing quantile regression curve estimation. Biometrika, 97, 825-838.
  4. [3] Chung, Y. and Dunson, D. B. (2009). Nonparametric Bayes conditional distribution modeling with variable selection. J. Amer. Statist. Assoc. 104, 1646-1660.
  5. [4] Dunson, D. B. and Taylor, J. A. (2005). Approximate Bayesian inference for quantiles. J. Nonparametri. Stat. 17, 385-400.
  6. [5] Feng, Y., Chen, Y., and He, X. (2015). Bayesian quantile regression with approximate likelihood. Bernoulli 21, 832-850.
  7. [6] He, X. (1997). Quantile curves without crossing. The American Statistician 51, 186-192.
  8. [7] He, X., Wang, L., and Hong, H.G. (2013). Quantile-adaptive model-free variable screening for high-dimensional heterogeneous data. Ann. Statist. 41, 342-369.
  9. [8] He, X. and Zhu, L. X. (2003). A lack-of-fit test for quantile regression. J. Amer. Statist. Assoc. 98, 1013-1022.
  10. [9] Jiang, L., Wang, H. J. and Bondell, H. D. (2013). Interquartile shrinkage in regression models. J. Comp. Graph. Statist., 22, 970-986.
  11. [10] Kai, B., Li, R. and Zou, H. (2011). New efficient estimation and variable selection methods for semiparametric varying-coefficient partially linear models. Ann. Statist. 39, 305-332.
  12. [11] Kato, K. (2011). Group Lasso for high dimensional sparse quantile regression models. arXiv preprint arXiv:1103.1458.
  13. [12] Kim, M. O. and Yang, Y. (2011). Semiparametric approach to a random effects quantile regression model. J. Amer. Statist. Assoc. 106, 1405-1417.
  14. [13] Koenker,
  15. R. (2005). Quantile Regression. Econometric Society Monographs 38. Cambridge Univ. Press, Cambridge.
  16. [14] Koenker, R. and Bassett, G. JR. (1978). Regression quantiles. Econometrica 46, 33–50.
  17. [15] Koenker, R. and Geling,
  18. O. (2001). Reappraising medfly longevity: A quantile regression survival analysis. J. Amer. Statist. Assoc. 96, 458-468.
  19. [16] Koenker, R. and Xiao, Z. (2002). Inference on the quantile regression process. Econometrica 70, 1583-1612.
  20. [17] M¨uller, P. and Quintana, F. A. (2004). Nonparametric Bayesian data analysis. Stat. Sci. 19, 95-110.
  21. [18] Newey, W. K. (1990). Semiparametric efficiency bounds. Journal of applied econometrics 5, 99-135.
  22. [19] Newey, W. K. and Powell, J. L. (1990). Efficient estimation of linear and type I censored regression models under conditional quantile restrictions. Econometric Theory 6, 295- 317.
  23. [20] Peng, L. and Fine, J. P. (2009). Competing risks quantile regression. J. Amer. Statist. Assoc. 104, 1440–1453.
  24. [21] Peng, L. and Huang, Y. (2008). Survival analysis with quantile regression models. J. Amer. Statist. Assoc. 103, 637-649.
  25. [22] Portnoy, S. (2003). Censored regression quantiles. J. Amer. Statist. Assoc. 98, 1001-1012.
  26. [23] Portnoy, S. (2012). Nearly root-n approximation for regression quantile processes. Ann. Statist 40, 1714-1736.
  27. [24] Portnoy, S. and Koenker, R. (1989). Adaptive L-estimation for linear models. Ann. Statist. 17, 362-381.
  28. [25] Portnoy, S. and Koenker, R. (1997). The Gaussian hare and the Laplacian tortoise: computability of squared-error versus absoluteerror estimators (with discussion). Statistical Science 12, 279-300.
  29. [26] Qu, Z. and Yoon, J. (2015). Nonparametric estimation and inference on conditional quantile processes. J. Econometrics 185, 1-19.
  30. [27] Reich, B. J., Fuentes, M., and Dunson, D. B. (2011). Bayesian spatial quantile regression. J. Amer. Statist. Assoc. 106, 6-20.
  31. [28] Tsiatis
  32. A. (2007). Semiparametric Theory and Missing Data. Springer Science & Business Media.
  33. [29] Wang, H. J. and Li, D. (2013). Estimation of Extreme Conditional Quantiles Through Power Transformation. J. Amer. Statist. Assoc. 108, 1062-1074.
  34. [30] Wang, H. J., Li, D. and He, X. (2012). Estimation of high conditional quantiles for heavy-tailed distributions. J. Amer. Statist. Assoc. 107, 1453-1464.
  35. [31] Wang, H. J. and Wang, L. (2009). Locally weighted censored quantile regression. J. Amer. Statist. Assoc. 104, 1117–1128.
  36. [32] Wang, L., Wu, Y., and Li, R. (2012). Quantile regression for analyzing heterogeneity in ultra-high dimension. J. Amer. Statist. Assoc. 107, 214-222.
  37. [33] Yang, Y. and He, X. (2012). Bayesian empirical likelihood for quantile regression. Ann. Statist. 40, 1102-1131.
  38. [34] Yu, K. and Jones, M. C. (1998). Local linear quantile regression. J. Amer. Statist. Assoc. 93, 228-237.
  39. [35] Zheng, Q., Peng, L., and He, X. (2015). Globally adaptive quantile regression with ultra-high dimensional data. Ann. Statist. 43, 22252258.
  40. [36] Zhou, K. Q. and Portnoy, S. L. (1998). Statistical inference on heteroscedastic models based on regression quantiles. J. Nonparametric Stat.9, 239-260.
  41. [37] Zou, H. and Yuan, M. (2008). Composite quantile regression and the oracle model selection theory. Ann. Statist. 36, 1108-1126.

Acknowledgments

The authors thank the Editor, the Associate Editor and two anonymous

reviewers for their insightful comments and constructive suggestions that

helped improve the paper significantly. Z. Wang’s research was supported

by the National Science Foundation of China (No. 12371277, No. 12231017).

Y. Lin’s research was partially supported by the Hong Kong Research

Grants Council (No. 14306620 and 14304523), and Direct Grants for Research, The Chinese University of Hong Kong.