On pointwise adaptive curve estimation based on inhomogeneous data

Stéphane Gaïffas

ESAIM: Probability and Statistics (2007)

  • Volume: 11, page 344-364
  • ISSN: 1292-8100

Abstract

top
We want to recover a signal based on noisy inhomogeneous data (the amount of data can vary strongly on the estimation domain). We model the data using nonparametric regression with random design, and we focus on the estimation of the regression at a fixed point x0 with little, or much data. We propose a method which adapts both to the local amount of data (the design density is unknown) and to the local smoothness of the regression function. The procedure consists of a local polynomial estimator with a Lepski type data-driven bandwidth selector, see for instance Lepski et al. [Ann. Statist.25 (1997) 929–947]. We assess this procedure in the minimax setup, over a class of function with local smoothness s > 0 of Hölder type. We quantify the amount of data at x0 in terms of a local property on the design density called regular variation, which allows situations with strong variations in the concentration of the observations. Moreover, the optimality of the procedure is proved within this framework.

How to cite

top

Gaïffas, Stéphane. "On pointwise adaptive curve estimation based on inhomogeneous data." ESAIM: Probability and Statistics 11 (2007): 344-364. <http://eudml.org/doc/250082>.

@article{Gaïffas2007,
abstract = { We want to recover a signal based on noisy inhomogeneous data (the amount of data can vary strongly on the estimation domain). We model the data using nonparametric regression with random design, and we focus on the estimation of the regression at a fixed point x0 with little, or much data. We propose a method which adapts both to the local amount of data (the design density is unknown) and to the local smoothness of the regression function. The procedure consists of a local polynomial estimator with a Lepski type data-driven bandwidth selector, see for instance Lepski et al. [Ann. Statist.25 (1997) 929–947]. We assess this procedure in the minimax setup, over a class of function with local smoothness s > 0 of Hölder type. We quantify the amount of data at x0 in terms of a local property on the design density called regular variation, which allows situations with strong variations in the concentration of the observations. Moreover, the optimality of the procedure is proved within this framework. },
author = {Gaïffas, Stéphane},
journal = {ESAIM: Probability and Statistics},
keywords = {Adaptive estimation; inhomogeneous data; nonparametric regression; random design.; adaptive estimation; random design},
language = {eng},
month = {8},
pages = {344-364},
publisher = {EDP Sciences},
title = {On pointwise adaptive curve estimation based on inhomogeneous data},
url = {http://eudml.org/doc/250082},
volume = {11},
year = {2007},
}

TY - JOUR
AU - Gaïffas, Stéphane
TI - On pointwise adaptive curve estimation based on inhomogeneous data
JO - ESAIM: Probability and Statistics
DA - 2007/8//
PB - EDP Sciences
VL - 11
SP - 344
EP - 364
AB - We want to recover a signal based on noisy inhomogeneous data (the amount of data can vary strongly on the estimation domain). We model the data using nonparametric regression with random design, and we focus on the estimation of the regression at a fixed point x0 with little, or much data. We propose a method which adapts both to the local amount of data (the design density is unknown) and to the local smoothness of the regression function. The procedure consists of a local polynomial estimator with a Lepski type data-driven bandwidth selector, see for instance Lepski et al. [Ann. Statist.25 (1997) 929–947]. We assess this procedure in the minimax setup, over a class of function with local smoothness s > 0 of Hölder type. We quantify the amount of data at x0 in terms of a local property on the design density called regular variation, which allows situations with strong variations in the concentration of the observations. Moreover, the optimality of the procedure is proved within this framework.
LA - eng
KW - Adaptive estimation; inhomogeneous data; nonparametric regression; random design.; adaptive estimation; random design
UR - http://eudml.org/doc/250082
ER -

References

top
  1. A. Antoniadis, G. Gregoire and P. Vial, Random design wavelet curve smoothing. Statist. Probab. Lett.35 (1997) 225–232.  
  2. Y. Baraud, Model selection for regression on a random design. ESAIM Probab. Statist.6 (2002) 127–146 (electronic).  
  3. N.H. Bingham, C.M. Goldie and J.L. Teugels, Regular Variation. Encyclopedia of Mathematics and its Applications, Cambridge University Press (1989).  
  4. L. Brown and T. Cai, Wavelet shrinkage for nonequispaced samples. Ann. Statist.26 (1998) 1783–1799.  
  5. L.D. Brown and M.G. Low, A constrained risk inequality with applications to nonparametric functional estimations. Ann. Statist.24 (1996) 2524–2535.  
  6. T.T. Cai, M. Low and L.H. Zhao, Tradeoffs between global and local risks in nonparametric function estimation. Tech. rep., Wharton, University of Pennsylvania, (2004).  URIhttp://stat.wharton.upenn.edu/~tcai/paper/html/Tradeoff.html
  7. V. Delouille, J. Simoens and R. Von Sachs, Smooth design-adapted wavelets for nonparametric stochastic regression. J. Amer. Statist. Soc.99 (2004) 643–658.  
  8. J. Fan and I. Gijbels, Data-driven bandwidth selection in local polynomial fitting: variable bandwidth and spatial adaptation. J. Roy. Statist. Soc. Ser. B. Methodological57 (1995) 371–394.  
  9. J. Fan and I. Gijbels, Local polynomial modelling and its applications. Monographs on Statistics and Applied Probability, Chapman & Hall, London (1996).  
  10. S. Gaïffas, Convergence rates for pointwise curve estimation with a degenerate design. Mathematical Methods of Statistics1 (2005) 1–27. Available at  URIhttp://hal.ccsd.cnrs.fr/ccsd-00003086/en/
  11. A. Goldenshluger and A. Nemirovski, On spatially adaptive estimation of nonparametric regression. Math. Methods Statist.6 (1997) 135–170.  
  12. G. Kerkyacharian and D. Picard, Regression in random design and warped wavelets. Bernoulli, 10 (2004) 1053–1105.  
  13. O.V. Lepski, Asymptotically minimax adaptive estimation i: Upper bounds, optimally adaptive estimates. Theory Probab. Applic.36 (1988) 682–697.  
  14. O.V. Lepski, On a problem of adaptive estimation in Gaussian white noise. Theory Probab. Appl., 35 (1990) 454–466.  
  15. O.V. Lepski, E. Mammen and V.G. Spokoiny, Optimal spatial adaptation to inhomogeneous smoothness: an approach based on kernel estimates with variable bandwidth selectors. Ann. Statist.25 (1997) 929–947.  
  16. O.V. Lepski and V.G. Spokoiny, Optimal pointwise adaptive methods in nonparametric estimation. Ann. Statist.25 (1997) 2512–2546.  
  17. V. Maxim, Restauration de signaux bruités sur des plans d'experience aléatoires. Ph.D. thesis, Université Joseph Fourier, Grenoble 1 (2003).  
  18. V.G. Spokoiny, Estimation of a function with discontinuities via local polynomial fit with an adaptive window choice. Ann. Statist.26 (1998) 1356–1378.  
  19. C.J. Stone, Optimal rates of convergence for nonparametric estimators. Ann. Statist.8 (1980) 1348–1360.  
  20. A. Tsybakov, Introduction à l'estimation non-paramétrique. Springer (2003).  
  21. M.-Y. Wong and Z. Zheng, Wavelet threshold estimation of a regression function with random design. J. Multivariate Anal.80 (2002) 256–284.  

NotesEmbed ?

top

You must be logged in to post comments.

To embed these notes on your page include the following JavaScript code on your page where you want the notes to appear.

Only the controls for the widget will be shown in your chosen language. Notes will be shown in their authored language.

Tells the widget how many notes to show per page. You can cycle through additional notes using the next and previous controls.

    
                

                
Note: Best practice suggests putting the JavaScript code just before the closing </body> tag.