Science in China Series A: Mathematics

, Volume 43, Issue 1, pp 65–81 | Cite as

Variable bandwidth and one-step local M-estimator

  • Jianqing Fan
  • Jiancheng Jiang


A robust version of local linear regression smoothers augmented with variable bandwidth is studied. The proposed method inherits the advantages of local polynomial regression and overcomes the shortcoming of lack of robustness of leastsquares techniques. The use of variable bandwidth enhances the flexibility of the resulting local M-estimators and makes them possible to cope well with spatially inhomogeneous curves, heteroscedastic errors and nonuniform design densities. Under appropriate regularity conditions, it is shown that the proposed estimators exist and are asymptotically normal. Based on the robust estimation equation, one-step local M-estimators are introduced to reduce computational burden. It is demonstrated that the one-step local M-estimators share the same asymptotic distributions as the fully iterative M-estimators, as long as the initial estimators are good enough. In other words, the onestep local M-estimators reduce significantly the computation cost of the fully iterative M-estimators without deteriorating their performance. This fact is also illustrated via simulations.


local regression M-estimator nonparametric estimation one-step robustness variable bandwidth 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Fan, J., Gijbels, I., Local Polynomical Modelling and Its Applications, London: Chapman and Hall, 1996.Google Scholar
  2. 2.
    Härdle, W., Applied Nonparametric Regression, New York: Cambridge University Press, 1990.MATHGoogle Scholar
  3. 3.
    Wand, M. P., Jones, M., Kernel Smoothing, London: Chapman and Hall, 1995.MATHGoogle Scholar
  4. 4.
    Fan, J., Local linear regression smoothers and their minimax efficiencies, Ann. Statist., 1993, 21: 196.MATHCrossRefMathSciNetGoogle Scholar
  5. 5.
    Ruppert, D., Wand, M. P., Multivariate weighted least squares regression, Ann. Statist., 1994, 22: 1–346CrossRefMathSciNetGoogle Scholar
  6. 6.
    Cleveland, W. S., Robust locally weighted regression and smoothing scatterplots, J. Amer. Statist. Assoc., 1979, 74: 829.MATHCrossRefMathSciNetGoogle Scholar
  7. 7.
    Cox, D. D., Asymptotics for M-type smoothing splines, Ann. Statist., 1983, 11: 530.MATHCrossRefMathSciNetGoogle Scholar
  8. 8.
    Tsybakov, A. B Robust reconstruction of functions by the local-approximation method, Problems of Information Transmission, 1986, 22: 133.MATHGoogle Scholar
  9. 9.
    Härdie, W., Tsybakov, A.B. Robust nonparametric regression with simultaneous scale curve estimation, Ann. Statist., 1988, 16: 120.CrossRefMathSciNetGoogle Scholar
  10. 10.
    Cunningham, J. K., Eubank, R. L., Hsing, T., M-type smoothing splines with auxiliary scale estimation, Comput. Statist & Data Anal., 1991, 11: 43.CrossRefMathSciNetGoogle Scholar
  11. 11.
    Fan, J., Hu, T., Truong, Y., Robust non-parametric function estimation, Scand. J. Statist., 1994, 21: 433.MATHMathSciNetGoogle Scholar
  12. 12.
    Welsh, A. H., Robust estimation of smooth regression and spread functions and their derivatives, Statist. Sínica, 1996, 6: 347.MATHMathSciNetGoogle Scholar
  13. 13.
    Fan, J., Gijbels, I., Data-driven bandwidth selection in local polynomical fitting: variable bandwidth and spatial adaption. J. R. Statist., Soc. B, 1995, 57: 371.MATHMathSciNetGoogle Scholar
  14. 14.
    Bickel, P. J., One-step Huber estimates in linear models, J. Amer. Stat. Assoc., 1975, 70: 428.MATHCrossRefMathSciNetGoogle Scholar
  15. 15.
    Breiman, L., Meisel, W., Purcell, E., Variable kernel estimates of multivariate densities, Technometrics, 1977, 19: 135.MATHCrossRefGoogle Scholar
  16. 16.
    Müller, H. -G., Stadtmüller, U., Variable bandwidth kernel estimators of regression curves, Ann. Statist., 1987, 15: 182.MATHCrossRefMathSciNetGoogle Scholar
  17. 17.
    Hall, P., Hu, T.C., Marron, J. S., Improved variable window estimators of probability densities, Ann. Statist., 1995, 23: 1.MATHCrossRefMathSciNetGoogle Scholar
  18. 18.
    Fan, J., Gijbels, I., Variable bandwidth and local linear regression smoothers, Ann. Statist., 1992, 20: 2008.MATHCrossRefMathSciNetGoogle Scholar
  19. 19.
    Gasser, T., Müller, H. -G., Kernel estimation of regression functions, in Smoothing Techniques for Curve Estimation, Lecture Notes in Mathematics, New York: Springer-Verlag, 1979, 23–68.CrossRefGoogle Scholar

Copyright information

© Science in China Press 2000

Authors and Affiliations

  • Jianqing Fan
    • 1
  • Jiancheng Jiang
    • 2
  1. 1.University of North Carolina at Chapel Hill and Chinese University of Hong KongHong KongChina
  2. 2.Department of Probability & StatisticsPeking UniversityBeijingChina

Personalised recommendations