The Annals of Statistics

Pointwise and sup-norm sharp adaptive estimation of functions on the Sobolev classes

A. B. Tsybakov

Full-text: Open access

Abstract

The problem of nonparametric function estimation in the Gaussian white noise model is considered. It is assumed that the unknown function belongs to one of the Sobolev classes, with an unknown regularity parameter. Asymptotically exact adaptive estimators of functions are proposed on the scale of Sobolev classes, with respect to pointwise and sup-norm risks. It is shown that, unlike the case of $L_2$-risk, a loss of efficiency under adaptation is inevitable here. Bounds on the value of the loss of efficiency are obtained.

Article information

Source
Ann. Statist., Volume 26, Number 6 (1998), 2420-2469.

Dates
First available in Project Euclid: 21 June 2002

Permanent link to this document
https://projecteuclid.org/euclid.aos/1024691478

Digital Object Identifier
doi:10.1214/aos/1024691478

Mathematical Reviews number (MathSciNet)
MR1700239

Zentralblatt MATH identifier
0933.62028

Subjects
Primary: 62G05: Estimation 62G20: Asymptotic properties

Keywords
Gaussian white noise adaptive nonparametric estimation Sobolev class loss of efficiency under adaptation minimax risk exact constants

Citation

Tsybakov, A. B. Pointwise and sup-norm sharp adaptive estimation of functions on the Sobolev classes. Ann. Statist. 26 (1998), no. 6, 2420--2469. doi:10.1214/aos/1024691478. https://projecteuclid.org/euclid.aos/1024691478


Export citation

References

  • BARRON, A., BIRGE, L. and MASSART, P. 1995. Risk bounds for model selection via penalization. Technical Report 95.54, Univ. Paris-Sud. Z.
  • BIRGE, L. 1983. Approximation dans les espaces metriques et theorie de l'estimation. Z. ´ ´ Wahrsch. Verw. Gebiete 65 181 237. Z.
  • BIRGE, L. and MASSART, P. 1997. From model selection to adaptive estimation. In Festschrift for Z. Lucien Le Cam D. Pollard, E. Torgersen, G. L. Yang, eds. 55 87. Springer, New York. Z.
  • BROWN, L. and LOW, M. 1996. A constrained risk inequality with applications to nonparametric estimation. Ann. Statist. 24 2524 2535. Z.
  • BUTUCEA, C. 1998. Estimation non parametrique adaptative de la densite de probabilite: ´ ´ ´ vitesses de convergence et constante exacte. Ph.D. thesis, Univ. Paris VI. Z.
  • DONOHO, D. L. 1994. Asy mptotic minimax risk for sup-norm loss: solution via optimal recovery. Probab. Theory Related Fields 99 145 170. Z.
  • DONOHO, D. L. and JOHNSTONE, I. M. 1995. Adapting to unknown smoothness via wavelet shrinkage. J. Amer. Statist. Assoc. 90 1200 1224. Z.
  • DONOHO, D. L., JOHNSTONE, I. M., KERKy ACHARIAN, G. and PICARD, D. 1995. Wavelet shrinkage Z. asy mptopia? with discussion J. Roy. Statist. Soc. Ser. B 57 301 369. Z.
  • DONOHO, D. L. and LOW, M. 1992. Renormalization exponents and optimal pointwise rates of convergence. Ann. Statist. 20 944 970. Z.
  • EFROIMOVICH, S. YU. and PINSKER, M. S. 1984. Learning algorithm for nonparametric filtering. Automat. Remote Control 11 1434 1440. Z.
  • GOLDENSHLUGER, A. and NEMIROVSKII, A. 1997. On spatial adaptive estimation of nonparametric regression. Math. Methods Statist. 6 135 170. Z.
  • GOLUBEV, G. K. 1990. Quasi-linear estimates of signals in L. Problems Inform. Transmission 2 26 15 20. Z.
  • GOLUBEV, G. K. and NUSSBAUM, M. 1992. Adaptive spline estimates for nonparametric regression models. Theory Probab. Appl. 37 521 529. Z.
  • HARDLE, W., KERKy ACHARIAN, G., PICARD, D. and TSy BAKOV, A. B. 1998. Wavelets, Approxima¨ tion and Statistical Applications. Lecture Notes in Statist. 129. Springer, New York. Z.
  • IBRAGIMOV, I. A. and HASMINSKII, R. Z. 1981. Statistical Estimation: Asy mptotic Theory. Z. Springer, New York. Originally published in Russian in 1979. Z.
  • IBRAGIMOV, I. A. and HASMINSKII, R. Z. 1984. Nonparametric estimation of the value of a linear functional in Gaussian white noise. Theory Probab. Appl. 29 18 32. Z.
  • JUDITSKY, A. 1997. Wavelet estimators: adapting to unknown smoothness. Math. Methods Statist. 6 1 25. Z.
  • KNEIP, A. 1994. Ordered linear smoothers. Ann. Statist. 22 835 866. Z.
  • KOROSTELEV, A. P. 1993. Exact asy mptotically minimax estimator for nonparametric regression in uniform norm. Theory Probab. Appl. 38 737 743. Z.
  • LEADBETTER, M. R., LINDGREN, G. and ROOTZEN, H. 1986. Extremes and Related Properties of Random Sequences and Processes. Springer, New York. Z.
  • LEPSKI, O. V. 1990. One problem of adaptive estimation in Gaussian white noise. Theory Probab. Appl. 35 454 466. Z.
  • LEPSKI, O. V. 1991. Asy mptotically minimax adaptive estimation I: upper bounds. Optimal adaptive estimates. Theory Probab. Appl. 36 682 697. Z.
  • LEPSKI, O. V. 1992a. Asy mptotically minimax adaptive estimation II: statistical models without optimal adaptation. Adaptive estimates. Theory Probab. Appl. 37 433 468. Z.
  • LEPSKI, O. V. 1992b. On problems of adaptive estimation in white Gaussian noise. In Topics in Z. Nonparametric Estimation. R. Z. Khasminskii, ed. 87 106. Amer. Math. Soc., Providence, RI.
  • LEPSKI, O. V. 1996. Personal communication. Z.
  • LEPSKI, O. V., MAMMEN, E. and SPOKOINY, V. G. 1997. Ideal spatial adaptation to inhomogeneous smoothness: an approach based on kernel estimates with variable bandwidth selection. Ann. Statist. 25 929 947. Z.
  • LEPSKI, O. V. and SPOKOINY, V. G. 1997. Optimal pointwise adaptive methods in nonparametric estimation. Ann. Statist. 25 2512 2546. Z.
  • NEMIROVSKII, A. 1985. On nonparametric estimation of smooth regression functions. Soviet J. Comput. Sy st. Sci. 23 1 11. Z.
  • OUDSHOORN, K. 1996. Optimality and adaptivity in nonparametric regression. Ph.D. thesis, Univ. Utrecht. Z.
  • PINSKER, M. S. 1980. Optimal filtering of square integrable signals in Gaussian white noise. Problems Inform. Transmission 16 120 133. Z.
  • STONE, C. J. 1980. Optimal rates of convergence for nonparametric estimators. Ann. Statist. 8 1348 1360. Z.
  • STONE, C. J. 1982. Optimal global rates of convergence for nonparametric regression. Ann. Statist. 10 1040 1053. Z.
  • TSy BAKOV, A. B. 1997. Asy mptotically efficient signal estimation in L with general loss 2 functions. Problems Inform. Transmission 33 78 88.