Kybernetika 47 no. 6, 855-865, 2011

One Bootstrap suffices to generate sharp uniform Bounds in Functional Estimation

Paul Deheuvels

Abstract:

We consider, in the framework of multidimensional observations, nonparametric functional estimators, which include, as special cases, the Akaike-Parzen-Rosenblatt kernel density estimators ([1], [18], [20]), and the Nadaraya-Watson kernel regression estimators ([16], [22]). We evaluate the sup-norm, over a given set <span class="tex"><b>I</b></span>, of the difference between the estimator and a non-random functional centering factor (which reduces to the estimator mean for kernel density estimation). We show that, under suitable general conditions, this random quantity is consistently estimated by the sup-norm over <span class="tex"><b>I</b></span> of the difference between the original estimator and a bootstrapped version of this estimator. This provides a simple and flexible way to evaluate the estimator accuracy, through a single bootstrap. The present work generalizes former results of Deheuvels and Derzko [4], given in the setup of density estimation in <b>R</b>.

Keywords:

density estimation, nonparametric functional estimation, regression estimation, bootstrap, resampling methods, confidence regions, empirical processes

Classification:

62G05, 62G08, 62G09, 62G15, 62G20, 62G30

References:

  1. H. Akaike: An approximation of the density function. Ann. Inst. Statist. Math. 6 (1954), 127-132.   CrossRef
  2. P. J. Bickel and M. Rosenblatt: On some global measures of the deviations of density functions estimates. Ann. Statist. 1 (1973), 1071-1095.   CrossRef
  3. G. Claeskens and I. van Keilegom: Bootstrap confidence bands for regression curves and their derivatives. Ann. Statist. 31 (2003), 1852-1884.   CrossRef
  4. P. Deheuvels and G. Derzko: Asymptotic certainty bands for kernel density estimators based upon a bootstrap resampling scheme. In: Statistical Models and Methods for Biomedical and Technical systems, Stat. Ind. Technol., Birkhäuser, Boston MA 2008, pp. 171-186.   CrossRef
  5. P. Deheuvels and D. M. Mason: General asymptotic confidence bands based on kernel-type function estimators. Statist. Infer. Stoch. Processes 7 (2004), 225-277.   CrossRef
  6. R. M. Dudley: Uniform Central Limit Theorems. Cambridge University Press, Cambridge 1999.   CrossRef
  7. B. Efron: Bootstrap methods: Another look at the jackknife. Ann. Statist. 7 (1979), 1-26.   CrossRef
  8. U. Einmahl and D. M. Mason: Uniform in bandwidth consistency of kernel-type function estimators. Ann. Statist. 33 (2005), 1380-1403.   CrossRef
  9. E. Giné, V. Koltchinskii and L. Sakhanenko: Kernel density estimators: convergence in distribution for weighted sup-norms. Probab. Theory Related Fields 130 (2004), 167-198.   CrossRef
  10. E. Giné and D. M. Mason: On local $U$-statistic processes and the estimation of densities of functions of several sample variables. Ann. Statist. 35 (2007), 1105-1145.   CrossRef
  11. P. Hall: The Bootstrap and Edgeworth Expansion. Springer, New York 1992.   CrossRef
  12. W. Härdle and J. S. Marron: Bootstrap simultaneous error bars for nonparametric regression. Ann. Statist. 19 (1991), 778-796.   CrossRef
  13. W. Härdle and A. W. Bowman: Bootstrapping in nonparametric regression: Local adaptative smoothing and confidence bands. J. Amer. Statist. Assoc. 83 (1988), 102-110.   CrossRef
  14. G. Li and S. Datta: A bootstrap approach to nonparametric regression for right censored data. Ann. Inst. Statist. Math. 53 (2001), 708-729.   CrossRef
  15. D. M. Mason and M. A. Newton: A rank statistics approach to the consistency of a general bootstrap. Ann. Statist. 20 (1992), 1611-1624.   CrossRef
  16. E. A. Nadaraya: On estimating regression. Theor. Probab. Appl. 9 (1964), 141-142.   CrossRef
  17. D. Nolan and D. Pollard: $U$-processes: Rates of convergence. Ann. Statist. 15 (1987), 780-799.   CrossRef
  18. E. Parzen: On the estimation of probability density and mode. Ann. Math. Statist. 33 (1962), 1065-1076.   CrossRef
  19. D. Pollard: Convergence of Stochastic Processes. Springer, New York 1984.   CrossRef
  20. M. Rosenblatt: Remarks on some nonparametric estimates of a density function. Ann. Math. Statist. 27 (1956), 832-837.   CrossRef
  21. A. W. van der Vaart and J. A. Wellner: Weak Convergence and Empirical Processes with Aplications to Statistics. Springer, New York 1996.   CrossRef
  22. G. S. Watson: Smooth regression analysis. Sankhy\={a} A 26 (1964), 359-372.   CrossRef