Abstract
M-cross-validation criterion is proposed for selecting a smoothing parameter in a nonparametric median regression model in which a uniform weak convergency rate for the M-cross-validated local median estimate, and the upper and lower bounds of the smoothing parameter selected by the proposed criterion are established. The main contribution of this study shows a drastic difference from those encountered in the classical L 2–, L 1- cross-validation technique, which leads only to the consistency in the sense of the average. Obviously, our results are novel and nontrivial from the point of view of mathematics and statistics, which provides insight and possibility for practitioners substituting maximum deviation for average deviation to evaluate the performance of the data-driven technique.
Similar content being viewed by others
References
Härdle, W.: Applied Nonparametric Regression, Cambridge University Press, New York, 1990
Wahba, G.: Spline Models for Observational Data, SIAM, Philadelphia, 1990
Fan, J., Gijbels, I.: Local Polynomial Modeling and Its Applications, Chapman and Hall, London, 1996
Koenker, R., Bassett, G. S.: Regression quantiles. Econometrica, 46, 33–50 (1978)
Chaudhuri, P.: Nonparametric estimates of regression quantiles and their local Bahadur representation. Ann. Statist., 19, 760–777 (1991)
Chaudhuri, P.: Global nonparametric estimation of conditional quantile functions and their derivatives. J. Multivariate Anal., 39, 246–269 (1991)
Yu, K. M., Jones, M. C.: Local linear quantile regression. J. Amer. Statist. Assoc., 93, 228–237 (1998)
Cheng, K. F.: Nonparametric estimators for percentile regression functions. Comm. Statist. A, Theory Methods, 12, 681–692 (1983)
Truong, Y. K. N.: Asymptotic properties of kernel estimators based on local medians. Ann. Statist, 17, 606–617 (1989)
Bhattacharya, P. K., Gangopadhyay, A. K.: Kernel and nearest–neighbor estimation of a conditional quantile. Ann. Statist, 18, 1400–1415 (1990)
Gangopadhyay, A. K., Sen P. K.: Bootstrap confidence intervals for conditional quantile functions. Sankhy, Ser. A, 52, 346–363 (1990)
Honda, T.: Nonparametric estimation of a conditional quantile for alpha–mixing processes. Ann. Inst. Statist. Math., 52, 459–470 (2000)
Marron, J. S.: What does optimal bandwidth selection means for nonparametric regression estimation? In: Statistical Data Analysis Based on the L 1– Norm and Related Methods (Edited by Y. Dodge), 379–391 North Holland, Amsterdam, 1987
Marron, J. S.: Automatic smoothing parameter selection, a survey. Empirical Econom., 13, 187–208 (1989)
Härdle, W., Chen, R.: Nonparametric time series analysis, a selective review with examples, In: Proceedings of the 50th ISI Session (Beijing, 1995), 1, 375–394
Simonoff, J. S.: Smoothing Methods in Statistics, Springer–Verlag, New York, 1996
Yu, K. M.: Smoothing regression quantile by combining k–NN estimation with local linear kernel fitting. Statist. Sinica, 9, 759–774 (1999)
Yang, Y.: Median cross–validation criterion. Chinese Sci. Bull., 42, 1956–1958 (1997)
Yang, Y., Zheng, Z. G.: Asymptotic properties for median cross–validated nearest neighbor median estimate in nonparametric regression. Sci. China Ser. A, 40, 585–597 (1997)
Zheng, Z. G., Yang, Y.: cross–validation and median criterion. Statist. Sinica, 8, 907–921 (1998)
Ronchetti, E.: Robust model selection in regression. Statist. Probab. Lett., 3, 21–23 (1985)
Ronchetti, E., Staudte, R. G.: A robust version of Mallows’s C p , J. Amer. Statist. Assoc., 89, 550–559 (1994)
Wu, Y., Zen, M. M.: A strongly consistent information criterion for linear model selection based on Mestimation. Probability Theory and Related Fields, 113, 599–625 (1999)
Assaid, C. A., Birch, J. B.: Automatic bandwidth selection in robust nonparametric regression. J. Statist. Comput. Simulation, 66, 259–272 (2000)
Chen, X. R., Zhao, L. C.: M–Methods in Linear Model, Shanghai Scientific Technical Publishers, Shanghai, 1996
Härdle, W., Marron, J. S.: Optimal bandwidth selection in nonparametric regression function estimation. Ann. Statist., 13, 1465–1481 (1988)
Li, Q., Racine, J.: Cross–validated local linear nonparametric regression. Statist. Sinica, 14, 485–512 (2004)
Wong, W. H.: On the consistency of cross–validation in kernel nonparametric regression. Ann. Statist., 11, 1136–1141 (1983)
Li, K. C.: Consistency for cross–validated nearest neighbor estimates in nonparametric regression. Ann. Statist., 12, 230–240 (1984)
Yang, Y., Zheng, Z. G.: Asymptotic properties for cross–validated nearest neighbor median estimates in nonparametric regression, L 1–view, In: Probability and Statistics (Jiang, Z., Yan, S., Cheng, P. et al. eds.), 242–257, World Scientific, Singapore, 1992
Hampel, F. R., Ronchetti, E. M, Rousseeuw, P. J., Stahel, W. A.: Robust Statistics: The Approach Based on Influence Functions. John Wiley & Sons, New York, 1986
Reiss, R. D.: Approximate Distributions of Order Statistics, Springer–Verlag, New York, 1989
Härdle, W., Hall, P., Marron, J. S.: How far automatically chosen regression smoothing parameters from their optimum? J. Amer. Statist. Assoc., 83, 86–95 (1988)
Hart, J., Yi, S.: One–side cross–validation. J. Amer. Statist. Assoc., 93, 620–631 (1998)
Chu, C. K., Marron, J. S.: Comparison of two bandwidth selectors with dependent errors. Ann. Statist., 19, 1906–1918 (1991)
Serfling, R. J.: Approximation Theorem of Mathematical Statistics, John Wiley, New York, 1980
von Bahr, B., Esseen, C.,G.: Inequalities for the rth absolute moment of a sum of random variables 1 ≤ r ≤ 2. Ann. Math. Statist., 36, 299–303 (1965)
Cole, R. H.: Relations between moments of order statistics. Ann. Math. Stat., 22, 308–310 (1951)
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Yang, Y. M-Cross-Validation in Local Median Estimation. Acta Math Sinica 22, 1565–1582 (2006). https://doi.org/10.1007/s10114-005-0855-3
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10114-005-0855-3
Keywords
- local median estimate
- cross-validation
- nonparametric median regression
- smoothing parameter
- uniform weak convergency rate