Abstract
This paper deals with recovering an unknown vector β from the noisy data Y = Xβ + σξ, where X is a known n × p matrix with n ≥ p and ξ is a standard white Gaussian noise. In order to estimate β, a spectral cut-off estimate \(\hat \beta ^{\bar m} (Y)\) with a data-driven cut-off frequency \(\bar m(Y)\) is used. The cut-off frequency is selected as a minimizer of the unbiased risk estimate of the mean square prediction error, i.e., \(\bar m = \arg \min _m \left\{ {\left\| {Y - X\hat \beta ^m \left( Y \right)} \right\|^2 + 2\sigma ^2 m} \right\}\). Assuming that β belongs to an ellipsoid W, we derive upper bounds for the maximal risk \(\sup _{\beta \in \mathbb{W}} E\left\| {\hat \beta ^{\bar m} \left( Y \right) - \beta } \right\|^2\) and show that \(\hat \beta ^{\bar m} \left( Y \right)\) is a rate optimal minimax estimator overW.
Similar content being viewed by others
References
H. Akaike, “Information Theory and an Extension of the Maximum Likelihood Principle”, in Proc. 2nd Intern. Symp. Inf. Theory (1973), pp. 267–281.
D. L. Donoho, “Nonlinear Solutions of Linear Inverse Problems by Wavelet-Vaguelette Decomposition”, J. Appl. and Comp. Harmonic Anal. 2, 101–126 (1995).
D. L. Donoho and I. M. Johnstone, “Neoclassical Minimax Problems, Thresholding and Adaptive Function Estimation”, Bernoulli 2, 39–62 (1996).
L. Cavalier and Yu. Golubev, “Risk Hull Method and Regularization by Projections of Ill-Posed Inverse Problems”, Ann. Statist. 34(4), 1653–1677 (2006).
H. W. Engl, M. Hanke, and A. Neubauer, Regularization of Inverse Problems, in Math. and Its Appl. (Kluwer, Dordrecht, 1996), Vol. 375.
M. S. Ermakov, “Minimax Estimation of the Solution of an Ill-Posed Convolution Type Problem”, Probl. Inform. Trans. 25, 191–200 (1989).
G. H. Golub, M. Heath, and G. Wahba, “Generalized Cross-Validation as a Method to Choosing a Good Ridge Parameter”, Technometrics 21, 215–223 (1979).
G. H. Golub and U. von Matt, “Generalized Cross-Validation for Large Scale Problems”, in Recent Advances in Total Least Squares Techniques and Errors in Variable Modeling, Ed. by S. van Huffel (SIAM, 1997), pp. 139–148.
Yu. Golubev, “On Universal Oracle Inequalities Related to High Dimensional Linear Models”, Ann. Statist. 38, 2751–2780 (2010).
Yu. Golubev, “ExponentialWeighting and Oracle Inequalities for ProjectionEstimates”, Probl. Inform.Trans. 48, 269–280 (2012).
A. Kneip, “Ordered Linear Smoothers”, Ann. Statist. 22, 835–866 (1994).
B. Mair and F.H. Ruymgaart, “Statistical Estimation in Hilbert Scale”, SIAM J. Appl.Math. 56, 1424–1444 (1996).
C. L. Mallows, “Some Comments on Cp”, Technometrics 15, 661–675 (1973).
J. Y. Koo, “Optimal Rates of Convergence for Nonparametric Statistical Inverse Problems”, Ann. Statist. 21, 590–599 (1993).
M. S. Pinsker, “Optimal Filtering of Square Integrable Signals in Gaussian White Noise”, Probl. Inform. Trans. 16, 120–133 (1980).
A. N. Tikhonov and V. A. Arsenin, Solution of Ill-posed Problems, in Scripta Series in Mathematics, (Wiley, New York, 1977).
A. Van der Vaart and J. A. Wellner, Weak Convergence and Empirical Processes. (Springer, New York, 1996).
Author information
Authors and Affiliations
Corresponding author
About this article
Cite this article
Chernousova, E., Golubev, Y. Spectral cut-off regularizations for ill-posed linear models. Math. Meth. Stat. 23, 116–131 (2014). https://doi.org/10.3103/S1066530714020033
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.3103/S1066530714020033
Keywords
- ill-posed linear model
- spectral cut-off regularization
- data-driven cut-off frequency
- oracle inequality
- minimax risk