An Effective Nonparametric Quantile Regression Method for Solving the Crossing Problem in Data Fitting Process

Abstract:

Nonparametric quantile regression method can be used as the first choice for some biostatistical data. Since the nonparametric quantile regression curves are estimated individually, the quantile curves can cross, leading to an invalid distribution estimation for the response. A simple nonparametric quantile regression method is proposed to avoid the crossing problem. The method uses nonparametric conditional density function estimate instead of the conditional distribution estimate to assure quantile regression function monotonous. Both a simulation study and an analysis of real salmon lustrousness data show the significant improvement of the method in solving the quantile crossing problem for some kind of biological data.

• Periodical: Applied Mechanics and Materials (Volumes 631-632)
• Pages: 245-249
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.631-632.245
• Online since: September 2014
• Authors: Huan Wang*, Jian Yuan Huang, Yong Sheng Yuan
• Keywords: Conditional Density, Conditional Distribution, Crossing Problem, Nonparametric Quantile Regression

References

[1] R. Keonker and G. Bassett, Regression Quantiles, Ecnometrica. Vol. 46(1978), pp.33-50.

[2] K. Yu, Z. Lu, and J. Stander, Quantile regression applications and current research areas, The Statistican. Vol. 52 (2003), pp.331-350.

[3] R. Koenker, Quantile regression. Cambridge University Press , (2005), pp.125-134.

[4] K. Yu and M.C. Jones, A comparison of local constant and local linear regression quantile estimation, Computational statistics and data analysis. Vol. 25 (1997), pp.159-166.

DOI: 10.1016/s0167-9473(97)00006-6

[5] K. Yu and M. C. Jones, Local linear quantile regression, Journal of the American Statistical Association, Vol. 93 (1998), pp.228-237.

DOI: 10.1080/01621459.1998.10474104

[6] P. Hall, R.C.L. Wolff, and Q. Yao, Methods for estimating a conditional distribution, Journal of the American Statistical Association, Vol. 94(1999), pp.154-163.

DOI: 10.1080/01621459.1999.10473832

[7] J. Fan and Q. Yao, Nonlinear Time Series: Nonparametric and Parametric Methods, New York: Springer-Verlag , (2003), pp.56-75.

[8] J. Fan, T. H. Yim, A crossvalidation method for estimating conditional densities, Biometrika, Vol. 91 (2004), pp.819-834.

DOI: 10.1093/biomet/91.4.819

[9] P. Hall, J. Racine, and Q. Li, Cross-validation and the estimation of conditional probability densities, Journal of the American Statistical Association, Vol. 99 (2004), pp.1015-1026.

DOI: 10.1198/016214504000000548

[10] R. J. Hyndman, D. M. Bashtannyk and G. K. Grunwald, Estimating and visualizing conditional densities, J. Comp. Graph. Statist, Vol. 5(1996), p.315–336.

DOI: 10.1080/10618600.1996.10474715

[11] D. M. Bastannyk and R. J. Hydman, Bandwidth selection for kernel conditional density estimation, Computational statistics and data analysis, Vol. 36 (2001), p.279–298.

DOI: 10.1016/s0167-9473(00)00046-3

[12] M. C. Jones, J.S. Marron, and S.J. Sheather, A brief survey of bandwith selection for density estimation, Journal of the American Statistical Association, Vol. 91(1996), pp.401-407.

DOI: 10.1080/01621459.1996.10476701

[13] X. Wang, Nonparametric statistics, Beijing: China Renmin University Press, Beijing, Beijing, (2000), pp.79-101.