The Annals of Statistics

On bandwidth choice in nonparametric regression with both short- and long-range dependent errors

Peter Hall, Soumendra Nath Lahiri, and Jörg Polzehl

Full-text: Open access

Abstract

We analyse methods based on the block bootstrap and leave-out cross-validation, for choosing the bandwidth in nonparametric regression when errors have an almost arbitrarily long range of dependence. A novel analytical device for modelling the dependence structure of errors is introduced. This allows a concise theoretical description of the way in which the range of dependence affects optimal bandwidth choice. It is shown that, provided block length or leave-out number, respectively, are chosen appropriately, both techniques produce first-order optimal bandwidths. Nevertheless, the block bootstrap has far better empirical properties, particularly under long-range dependence.

Article information

Source
Ann. Statist., Volume 23, Number 6 (1995), 1921-1936.

Dates
First available in Project Euclid: 15 October 2002

Permanent link to this document
https://projecteuclid.org/euclid.aos/1034713640

Digital Object Identifier
doi:10.1214/aos/1034713640

Mathematical Reviews number (MathSciNet)
MR1389858

Zentralblatt MATH identifier
0856.62041

Subjects
Primary: 62G07: Density estimation 62G09: Resampling methods
Secondary: 62M10: Time series, auto-correlation, regression, etc. [See also 91B84]

Keywords
Bandwidth choice block bootstrap correlated errors cross-validation curve estimation kernel estimator local linear smoothing long-range dependence mean squared error nonparametric regression resampling short-range dependence

Citation

Hall, Peter; Lahiri, Soumendra Nath; Polzehl, Jörg. On bandwidth choice in nonparametric regression with both short- and long-range dependent errors. Ann. Statist. 23 (1995), no. 6, 1921--1936. doi:10.1214/aos/1034713640. https://projecteuclid.org/euclid.aos/1034713640


Export citation

References

  • ALTMAN, N. S. 1990. Kernel smoothing of data with correlated errors. J. Amer. Statist. Assoc. 85 749 759. Z.
  • CARLSTEIN, E. 1986. The use of subseries methods for estimating the variance of a general statistic from a stationary time series. Ann. Statist. 14 1171 1179. Z.
  • CHIU, S.-T. 1989. Bandwidth selection for kernel estimation with correlated noise. Statist. Probab. Lett. 8 347 354. Z.
  • CHU, C. K. and MARRON, J. S. 1991. Comparison of two bandwidth selectors with dependent errors. Ann. Statist. 19 1906 1918. Z.
  • DIGGLE, P. J. and HUTCHINSON, M. F. 1989. On spline smoothing with autocorrelated errors. Austral. J. Statist. 31 166 182. Z.
  • FAN, J. 1993. Local linear smoothers and their minimax efficiency. Ann. Statist. 21 196 216. Z.
  • FARAWAY, J. J. and JHUN, M. 1990. Bootstrap choice of bandwidth for density estimation. J. Amer. Statist. Assoc. 85 1119 1122. Z.
  • GASSER, T and MULLER, H. G. 1979. Kernel estimation of regression functions. Smoothing ¨ Techniques for Curve Estimation Lecture Notes in Math. 757 23 68. Springer, Berlin. Z.
  • GASSER, T., ENGLE, J. and SEIFERT, B. 1993. Nonparametric function estimation. In Handbook Z. of Statistics C. R. Rao, ed. 9 423 465. North-Holland, Amsterdam. Z.
  • Gy ORFI, L., HARDLE, W., SARDA, P. and VIEU, P. 1989. Nonparametric Curve Estimation from ¨ ¨ Time Series. Lecture Notes in Statist. 60. Springer, New York. Z.
  • HALL, P. 1985. Resampling a coverage pattern. Stochastic Process. Appl. 20 231 246. Z.
  • HALL, P. and HEy DE, C. C. 1980. Martingale Limit Theory and Its Application. Academic Press, New York. Z.
  • HARDLE, W. 1990. Applied Nonparametric Regression. Cambridge Univ. Press. ¨ Z.
  • HARDLE, W. and VIEU, P. 1992. Kernel regression smoothing of time series. J. Time Ser. Anal. ¨ 13 209 232. Z.
  • HART, J. D. 1987. Kernel smoothing when the observations are correlated. Technical report, Dept. Statistics, Texas A & M Univ. Z.
  • HART, J. D. 1991. Kernel regression estimation with time series errors. J. Roy. Statist. Soc. Ser. B 53 173 187. Z.
  • HART, J. D. 1994. Automated kernel smoothing of dependent data by using time series cross-validation. J. Roy. Statist. Soc. Ser. B 56 529 542. Z.
  • HART, J. D. and VIEU, P. 1990. Data-driven bandwidth choice for density estimation based on dependent data. Ann. Statist. 18 873 890. Z.
  • HART, J. D. and WEHRLY, T. E. 1986. Kernel regression using repeated measurements data. J. Amer. Statist. Assoc. 81 1080 1088. Z.
  • HERMANN, E., GASSER, T. and KNEIP, A. 1992. Choice of bandwidth for kernel regression when residuals are correlated. Biometrika 79 783 795. Z.
  • KOHN, R., ANSLEY, C. G. and WONG, C.-M. 1992. Nonparametric spline regression with autoregressive moving average errors. Biometrika 79 335 346. Z.
  • KUNSCH, H. R. 1989. The jackknife and the bootstrap for general stationary observations. Ann. ¨ Statist. 17 1217 1241.
  • TAy LOR, C. C. 1989. Bootstrap choice of the smoothing parameter in kernel density estimation. Biometrika 76 705 712. Z. d
  • WOOD, A. T. A. and CHAN, G. 1993. Simulation of stationary Gaussian processes in 0, 1. Research Report CMA-SR06-93, Centre for Mathematics and its Applications, Australian National Univ.
  • AUSTRALIAN NATIONAL UNIVERSITY AMES, IOWA 80011 G.P.O. BOX 4
  • CANBERRA, ACT 0200 AUSTRALIA