We consider the estimation of quadratic functionals in a Gaussian sequence model where the eigenvalues are supposed to be unknown and accessible through noisy observations only. Imposing smoothness assumptions both on the signal and the sequence of eigenvalues, we develop a minimax theory for this problem. We propose a truncated series estimator and show that it attains the optimal rate of convergence if the truncation parameter is chosen appropriately. Consequences for testing problems in inverse problems are equally discussed: in particular, the minimax rates of testing for signal detection and goodness-of-fit testing are derived.
Accepté le :
DOI : 10.1051/ps/2018027
Mots-clés : Inverse problem, unknown eigenvalues, minimax theory, rate optimality, truncated series estimator, non-parametric testing, goodness-of-fit testing, signal detection
@article{PS_2019__23__524_0,
author = {Kroll, Martin},
title = {Rate optimal estimation of quadratic functionals in inverse problems with partially unknown operator and application to testing problems},
journal = {ESAIM: Probability and Statistics},
pages = {524--551},
publisher = {EDP-Sciences},
volume = {23},
year = {2019},
doi = {10.1051/ps/2018027},
mrnumber = {3990654},
zbl = {1422.62129},
language = {en},
url = {http://www.numdam.org/articles/10.1051/ps/2018027/}
}
TY - JOUR AU - Kroll, Martin TI - Rate optimal estimation of quadratic functionals in inverse problems with partially unknown operator and application to testing problems JO - ESAIM: Probability and Statistics PY - 2019 SP - 524 EP - 551 VL - 23 PB - EDP-Sciences UR - http://www.numdam.org/articles/10.1051/ps/2018027/ DO - 10.1051/ps/2018027 LA - en ID - PS_2019__23__524_0 ER -
%0 Journal Article %A Kroll, Martin %T Rate optimal estimation of quadratic functionals in inverse problems with partially unknown operator and application to testing problems %J ESAIM: Probability and Statistics %D 2019 %P 524-551 %V 23 %I EDP-Sciences %U http://www.numdam.org/articles/10.1051/ps/2018027/ %R 10.1051/ps/2018027 %G en %F PS_2019__23__524_0
Kroll, Martin. Rate optimal estimation of quadratic functionals in inverse problems with partially unknown operator and application to testing problems. ESAIM: Probability and Statistics, Tome 23 (2019), pp. 524-551. doi : 10.1051/ps/2018027. http://www.numdam.org/articles/10.1051/ps/2018027/
[1] and , Estimating integrated squared density derivatives: sharp best order of convergence estimates. Sankhyā Ser. A 50 (1988) 381–393. | MR | Zbl
[2] , , and , Testing for lack of fit in inverse regression—with applications to biophotonic imaging. J. R. Stat. Soc. Ser. B Stat. Methodol. 71 (2009) 25–48. | DOI | MR | Zbl
[3] , Goodness-of-fit testing and quadratic functional estimation from indirect observations. Ann. Stat. 35 (2007) 1907–1930. | DOI | MR | Zbl
[4] , and , Adaptive goodness-of-fit testing from indirect observations. Ann. Inst. Henri Poincaré Probab. Stat. 45 (2009) 352–372. | DOI | Numdam | MR | Zbl
[5] and , Quadratic functional estimation in inverse problems. Stat. Methodol. 8 (2011) 31–41. | DOI | MR | Zbl
[6] and , Nonquadratic estimators of a quadratic functional. Ann. Stat. 33 (2005) 2930–2956. | MR | Zbl
[7] and , Optimal adaptive estimation of a quadratic functional. Ann. Stat. 34 (2006) 2298–2325. | MR | Zbl
[8] , Inverse problems in statistics. In Inverse problems and high-dimensional estimation, Vol. 203 of Lect. Notes Stat. Proc. Springer, Heidelberg (2011) 3–96. | MR
[9] and , Adaptive estimation for inverse problems with noisy operators. Inverse Probl. 21 (2005) 1345–1361. | DOI | MR | Zbl
[10] , On adaptive wavelet estimation of a quadratic functional from a deconvolution problem. Ann. Inst. Stat. Math. 63 (2011) 405–429. | DOI | MR | Zbl
[11] , and , Minimax estimation of linear and quadratic functionals on sparsity classes. Ann. Stat. 45 (2017) 923–958. | DOI | MR | Zbl
[12] and , Data-driven density estimation in the presence of additive noise with unknown distribution. J. R. Stat. Soc. Ser. B Stat. Methodol. 73 (2011) 601–627. | DOI | MR | Zbl
[13] , Nonlinear solution of linear inverse problems by wavelet-vaguelette decomposition. Appl. Comput. Harmon. Anal. 2 (1995) 101–126. | DOI | MR | Zbl
[14] and , Minimax quadratic estimation of a quadratic functional. J. Complex. 6 (1990) 290–323. | DOI | MR | Zbl
[15] and , On inverse problems with unknown operators. IEEE Trans. Inf. Theory 47 (2001) 2876–2894. | DOI | MR | Zbl
[16] , Minimax detection of a signal in Gaussian white noise. Teor. Veroyatnost. i Primenen. 35 (1990) 704–715. | MR | Zbl
[17] , On the estimation of quadratic functionals. Ann. Stat. 19 (1991) 1273–1294. | MR | Zbl
[18] and , Minimax estimation of a bounded squared mean. Stat. Probab. Lett. 13 (1992) 383–390. | DOI | MR | Zbl
[19] and , Wavelet methods to estimate an integrated quadratic functional: adaptivity and asymptotic law. Stat. Probab. Lett. 44 (1999) 109–122. | DOI | MR | Zbl
[20] and , A simple adaptive estimator of the integrated square of a density. Bernoulli 14 (2008) 47–61. | DOI | MR | Zbl
[21] and , Nonlinear estimation for linear inverse problems with error in the operator. Ann. Stat. 36 (2008) 310–336. | DOI | MR | Zbl
[22] and , Nonparametric goodness-of-fit testing under Gaussian models. In Vol. 169 of Lecture Notes in Statistics. Springer-Verlag, New York (2003). | DOI | MR | Zbl
[23] , Asymptotically minimax hypothesis testing for nonparametric alternatives (i–iii). Math. Methods Stat. 2 (1993) 85–114, 171–189, 249–268. | MR | Zbl
[24] , and Minimax signal detection in ill-posed inverse problems. Ann. Stat. 40 (2012) 1524–1549. | DOI | MR | Zbl
[25] , Deconvolution with unknown error distribution. Ann. Stat. 37 (2009) 2301–2323. | DOI | MR | Zbl
[26] and , Adaptive Gaussian inverse regression with partially unknown operator. Commun. Stat. Theory Methods 42 (2013) 1343–1362. | DOI | MR | Zbl
[27] , Thresholding for weighted χ2. Stat. Sinica 11 (2001) 691–704. | MR | Zbl
[28] , Chi-square oracle inequalities. In State of the art in probability and statistics (Leiden, 1999). In Vol. 36 of IMS Lecture Notes Monogr. Ser. Inst. Math. Stat., Beachwood, OH (2001) 399–418. | DOI | MR
[29] , Sharp adaptive estimation of quadratic functionals. Probab. Theory Related Fields 134 (2006) 539–564. | DOI | MR | Zbl
[30] and , Goodness-of-fit test for noisy directional data. Bernoulli 20 (2014) 2131–2168. | DOI | MR | Zbl
[31] , and , Testing inverse problems: a direct or an indirect problem? J. Stat. Plann. Infer. 141 (2011) 1849–1861. | DOI | MR | Zbl
[32] and , Adaptive estimation of a quadratic functional by model selection. Ann. Stat. 28 (2000) 1302–1338. | DOI | MR | Zbl
[33] , Adaptive estimation of a quadratic functional of a density by model selection. ESAIM: PS 9 (2005) 1–18. | DOI | Numdam | MR | Zbl
[34] and , A unified treatment for non-asymptotic and asymptotic approaches to minimax signal detection. Stat. Surv. 9 (2015) 253–297. | DOI | MR | Zbl
[35] and , Minimax goodness-of-fit testing in ill-posed inverse problems with partially unknown operators. Ann. Inst. Henri Poincaré Probab. Stat. 53 (2017) 1675–1718. | DOI | MR | Zbl
[36] , On the effect of estimating the error density in nonparametric deconvolution. J. Nonparametr. Stat. 7 (1997) 307–330. | DOI | MR | Zbl
[37] and , The maxiset point of view for estimating integrated quadratic functionals. Stat. Sinica 18 (2008) 255–279. | MR | Zbl
[38] , Introduction to nonparametric estimation. In Springer Series in Statistics. Revised and extended fromthe 2004 French original, Translated by Vladimir Zaiats. Springer, New York (2009). | MR | Zbl
[39] , Aggregation and minimax optimality in high-dimensional estimation. In Proc. of the International Congress of Mathematicians—Seoul 2014. Vol. IV. Kyung Moon Sa, Seoul (2014) 225–246. | MR | Zbl
Cité par Sources :
