Bernoulli

  • Bernoulli
  • Volume 12, Number 4 (2006), 609-632.

Nonlinear estimation over weak Besov spaces and minimax Bayes method

Vincent Rivoirard

Full-text: Open access

Abstract

Weak Besov spaces play an important role in statistics as maxisets of classical procedures or for measuring the sparsity of signals. The goal of this paper is to study weak Besov balls WB s ,p,q (C) from the statistical point of view by using the minimax Bayes method. In particular, we compare weak and strong Besov balls statistically. By building an optimal Bayes wavelet thresholding rule, we first establish that, under suitable conditions, the rate of convergence of the minimax risk for WB s ,p,q (C) is the same as for the strong Besov ball B s ,p,q (C) that is contained in WB s ,p,q (C) . However, we show that the asymptotically least favourable priors of WB s ,p,q (C) that are based on Pareto distributions cannot be asymptotically least favourable priors for B s ,p,q (C) . Finally, we present sample paths of such priors that provide representations of the worst functions to be estimated for classical procedures and we give an interpretation of the roles of the parameters s , p and q of WB s ,p,q (C) .

Article information

Source
Bernoulli, Volume 12, Number 4 (2006), 609-632.

Dates
First available in Project Euclid: 16 August 2006

Permanent link to this document
https://projecteuclid.org/euclid.bj/1155735929

Digital Object Identifier
doi:10.3150/bj/1155735929

Mathematical Reviews number (MathSciNet)
MR2248230

Zentralblatt MATH identifier
1125.62001

Keywords
asymptotically least favourable priors Bayes method minimax risk rate of convergence thresholding rules weak Besov spaces

Citation

Rivoirard, Vincent. Nonlinear estimation over weak Besov spaces and minimax Bayes method. Bernoulli 12 (2006), no. 4, 609--632. doi:10.3150/bj/1155735929. https://projecteuclid.org/euclid.bj/1155735929


Export citation

References

  • [1] Abramovich, F., Benjamini, Y., Donoho, D.L. and Johnstone, I.M. (2000) Adapting to unknown sparsity by controlling the false discovery rate. Technical report.
  • [2] Abramovich, F., Sapatinas, T. and Silverman, B.W. (1998) Wavelet thresholding via a Bayesian approach. J. Roy. Statist. Soc. Ser. B, 60, 725-749.
  • [3] Antoniadis, A. (1994) Smoothing noisy data with coiflets. Statist. Sinica, 4, 651-678.
  • [4] Bickel, P.J. (1981) Minimax estimation of the mean of a normal distribution when the parameter space is restricted. Ann. Statist., 9, 1301-1309.
  • [5] Casella, G. and Strawderman, W.E. (1981) Estimating a bounded normal mean. Ann. Statist., 9, 870-878.
  • [6] Cohen, A., DeVore, R.A. and Hochmuth, R. (2000a) Restricted nonlinear approximation. Constr. Approx., 16, 85-113.
  • [7] Cohen, A., DeVore, R.A., Kerkyacharian, G. and Picard, D. (2000b) Maximal spaces with given rate of convergence for thresholding algorithms. Appl. Comput. Harmon. Anal., 11, 167-191.
  • [8] Daubechies, I. (1992) Ten Lectures on Wavelets. Philadelphia: SIAM.
  • [9] DeVore, R.A. (1989) Degree of nonlinear approximation. In C.K. Chui, L.L. Schumaker and J.D. Wards (eds), Approximation Theory VI, Vol. 1, pp. 175-201. New York: Academic Press.
  • [10] DeVore, R.A. and Lorentz, G.G. (1993) Constructive Approximation. New York: Springer-Verlag.
  • [11] Donoho, D.L. (1996) Unconditional bases and bit-level compression. Appl. Comput. Harmon. Anal., 3, 388-392.
  • [12] Donoho, D.L. and Johnstone, I.M. (1994a) Ideal spatial adaptation by wavelet shrinkage. Biometrika, 81, 425-455.
  • [13] Donoho, D.L. and Johnstone, I.M. (1994b) Minimax risk over l p-balls for lq-error. Probab. Theory Related Fields, 99, 277-303.
  • [14] Donoho, D.L. and Johnstone, I.M. (1996) Neo-classical minimax problems, thresholding and adaptive function estimation. Bernoulli, 2, 39-62.
  • [15] Donoho, D.L. and Johnstone, I.M. (1998) Minimax estimation via wavelet shrinkage. Ann. Statist., 26, 879-921.
  • [16] Donoho, D.L., Johnstone, I.M., Kerkyacharian, G. and Picard, D. (1996) Universal near minimaxity of wavelet shrinkage. In D. Pollard, E. Torgersen and G. Yangs (eds), Festschrift for Lucien Le Cam, pp. 183-218. New York: Springer-Verlag.
  • [17] Johnstone, I.M. (1994) Minimax Bayes, asymptotic minimax and sparse wavelet priors. In S.S. Gupta and J.O. Berger (eds), Statistical Decision Theory and Related Topics, Vol. 5, pp. 303-326. New York: Springer-Verlag.
  • [18] Mallat, S. (1998) A Wavelet Tour of Signal Processing. San Diego, CA: Academic Press.
  • [19] Massart, P. (1990) The tight constant in the Dvoretzky-Kiefer-Wolfowitz inequality. Ann. Probab., 18, 1269-1283.
  • [20] Meyer, Y. (1992) Wavelets and Operators. Cambridge: Cambridge University Press.
  • [21] Pinsker, M.S. (1980) Optimal filtering of square integrable signals in Gaussian white noise. Problems Inform. Transmission, 16, 120-133.
  • [22] Rivoirard, V. (2004) Thresholding procedure with priors based on Pareto distributions. Test, 13, 213-246.
  • [23] Rivoirard, V. (2005) Bayesian modelling of sparse sequences and maxisets for Bayes rules. Math. Methods Statist., 14, 346-376.
  • [24] Vidakovic, B. (1998) Nonlinear wavelet shrinkage with Bayes rules and Bayes factors. J. Amer. Statist. Assoc., 93, 173-179.