• Bernoulli
  • Volume 11, Number 2 (2005), 309-340.

Minimax estimation of the noise level and of the deconvolution density in a semiparametric convolution model

Cristina Butucea and Catherine Matias

Full-text: Open access


We consider a semiparametric convolution model where the noise has known Fourier transform which decays asymptotically as an exponential with unknown scale parameter; the deconvolution density is less smooth than the noise in the sense that the tails of the Fourier transform decay more slowly, ensuring the identifiability of the model. We construct a consistent estimation procedure for the noise level and prove that its rate is optimal in the minimax sense. Two convergence rates are distinguished according to different smoothness properties for the unknown density. If the tail of its Fourier transform does not decay faster than exponentially, the asymptotic optimal rate and exact constant are evaluated, while if it does not decay faster than polynomially, this rate is evaluated up to a constant. Moreover, we construct a consistent estimator of the unknown density, by using a plug-in method in the classical kernel estimation procedure. We establish that the rates of estimation of the deconvolution density are slower than in the case of an entirely known noise distribution. In fact, nonparametric rates of convergence are equal to the rate of estimation of the noise level, and we prove that these rates are minimax. In a few specific cases the plug-in method converges at even slower rates.

Article information

Bernoulli, Volume 11, Number 2 (2005), 309-340.

First available in Project Euclid: 17 May 2005

Permanent link to this document

Digital Object Identifier

Mathematical Reviews number (MathSciNet)

Zentralblatt MATH identifier

analytic densities deconvolution L_2 risk minimax estimation noise level pointwise risk semiparametric model Sobolev classes supersmooth densities


Butucea, Cristina; Matias, Catherine. Minimax estimation of the noise level and of the deconvolution density in a semiparametric convolution model. Bernoulli 11 (2005), no. 2, 309--340. doi:10.3150/bj/1116340297.

Export citation


  • [1] Butucea, C. (2001) Exact adaptive pointwise estimation on Sobolev classes of densities, ESAIM Probab. Statist., 5, 1-31.
  • [2] Butucea, C. and Tsybakov, A. (2004) Sharp optimality for density deconvolution with dominating bias. arXiv: math.ST/0409471.
  • [3] Carroll, R.J. and Hall, P. (1988) Optimal rates of convergence for deconvolving a density. J. Amer. Statist. Assoc., 83, 1184-1186.
  • [4] Efromovich, S. (1997) Density estimation for the case of supersmooth measurement error. J. Amer. Statist. Assoc., 92, 526-535.
  • [5] Goldenshluger, A. (1999) On pointwise adaptive nonparametric deconvolution, Bernoulli, 5, 907-925.
  • [6] Goldenshluger, A. (2002) Density deconvolution in the circular structural model. J. Multivariate Anal., 81, 360-375.
  • [7] Gradshteyn, I. and Ryzhik, I. (2000) Table of Integrals, Series, and Products, 6th edition. San Diego, CA: Academic Press.
  • [8] Koltchinskii, V.I. (2000) Empirical geometry of multivariate data: a deconvolution approach, Ann. Statist., 28, 591-629.
  • [9] Lindsay, B.G. (1986) Exponential family mixture models (with least-squares estimators). Ann. Statist., 14, 124-137.
  • [10] Matias, C. (2002) Semiparametric deconvolution with unknown noise variance, ESAIM Probab. Statist., 6, 271-292.
  • [11] van der Vaart, A.W. and Wellner, J.A. (1996) Weak Convergence and Empirical Processes. New York: Springer-Verlag.
  • [12] Zhang, C.-H. (1990) Fourier methods for estimating mixing densities and distributions, Ann. Statist., 18, 806-831.
  • [13] Zolotarev, V. (1986) One-Dimensional Stable Distributions, Transl. Math. Monogr. 65. Providence, RI: American Mathematical Society.