×
Laurent, B.; Massart, P.

Adaptive estimation of a quadratic functional by model selection. (English) Zbl 1105.62328

Ann. Stat. 28, No. 5, 1302-1338 (2000).
Summary: We consider the problem of estimating \(\| s\|^2\) when s belongs to some separable Hilbert space and one observes the Gaussian process \(Y(t) = \langle s, t\rangle + \sigma L(t)\), for all \(t \epsilon \mathbf H\),where \(L\) is some Gaussian isonormal process. This framework allows us in particular to consider the classical Gaussian sequence model for which \({\mathbf H} = l_2({\mathbf N}^*)\) and \(L(t) = \sum_{\lambda\geq1}t_{\lambda}\varepsilon_{\lambda}\), where \((\varepsilon_{\lambda})_{\lambda\geq1}\) is a sequence of i.i.d. standard normal variables. Our approach consists in considering some at most countable families of finite-dimensional linear subspaces of \(\mathbf H\) (the models) and then using model selection via some conveniently penalized least squares criterion to build new estimators of \(\| s\|^2\). We prove a general nonasymptotic risk bound which allows us to show that such penalized estimators are adaptive on a variety of collections of sets for the parameter s, depending on the family of models from which they are built.In particular, in the context of the Gaussian sequence model, a convenient choice of the family of models allows defining estimators which are adaptive over collections of hyperrectangles, ellipsoids, \(l_p\)-bodies or Besov bodies.We take special care to describe the conditions under which the penalized estimator is efficient when the level of noise \(\sigma\) tends to zero. Our construction is an alternative to the one by Efroïmovich and Low for hyperrectangles and provides new results otherwise.

Cited in 229 Documents

MSC:

62G05 Nonparametric estimation
46N30 Applications of functional analysis in probability theory and statistics
62M09 Non-Markovian processes: estimation
Cite Review PDF
Full Text: DOI

References:

[1] Baraud, Y. (2000). Model selection for regression on a fixed design. Probab. Theory Related Fields 117 467-493. · Zbl 0997.62027 · doi:10.1007/s004400000058
[2] Barron, A. R., Birgé, L. and Massart, P. (1999). Risk bound for model selection via penalization. Probab. Theory Related Fields 113 301-415. · Zbl 0946.62036 · doi:10.1007/s004400050210
[3] Bickel, P. and Ritov, Y. (1988). Estimating integrated squared density derivatives: sharp best order of convergence estimates. Sankhy \?a Ser. A 50 381-393. · Zbl 0676.62037
[4] Birgé, L. (1983). Approximation dans les espaces métriques et théorie de l’estimation. Z. Wahrsch. Verw. Gebiete 65 181-237. · Zbl 0506.62026 · doi:10.1007/BF00532480
[5] Birgé, L. and Massart, P. (1995). Estimation of integral functionals of a density. Ann. Statist. 23 11-29. · Zbl 0848.62022 · doi:10.1214/aos/1176324452
[6] Birgé, L. and Massart, P. (1997). From model selection to adaptive estimation. In Festschrift for Lucien Le Cam: Research Papers in Probability and Statistics (D. Pollard, E. Torgersen and G. Yang, eds.) 55-87. Springer, New York. · Zbl 0920.62042
[7] Birgé, L. and Massart, P. (1998). Minimum contrast estimators on sieves: exponential bounds and rates of convergence. Bernoulli 4 329-375. Birgé, L. and Massart, P. (2000a). An adaptive compression algorithm in Besov spaces. Constr. Approx. 16 1-36. Birgé, L. and Massart, P. (2000b). Gaussian model selection. Technical Report 2000.05, Univ. Paris Sud. · Zbl 0954.62033 · doi:10.2307/3318720
[8] DeVore, R. A., Jawerth, B. and Popov, V. (1992). Compression of wavelet decompositions. Amer. J. Math. 114 737-785. JSTOR: · Zbl 0764.41024 · doi:10.2307/2374796
[9] DeVore, R. A. and Lorentz, G. G. (1993). Constructive Approximation. Springer, New York. · Zbl 0797.41016
[10] DeVore, R. A., Kyriazis, G. Leviatan, D. and Tikhomirov, V. M. (1993). Wavelet compression and nonlinear n-widths. Adv. Comput. Math. 1 197-214. · Zbl 0824.65145 · doi:10.1007/BF02071385
[11] Johnstone, I. (1999). Chi-square oracle inequalities. · Zbl 1373.62062 · doi:10.1214/lnms/1215090080
[12] Donoho, D. and Johnstone, I. (1998). Minimax estimation via wavelet shrinkage. Ann. Statist. 26 879-921. · Zbl 0935.62041 · doi:10.1214/aos/1024691081
[13] Donoho, D. and Liu, R. (1991). Geometrizing rates of convergence II. Ann. Statist. 19 633-668. · Zbl 0754.62028 · doi:10.1214/aos/1176348114
[14] Donoho, D. and Nussbaum, M. (1990). Minimax quadratic estimation of a quadratic functional. J. Complexity 6 290-323. · Zbl 0724.62039 · doi:10.1016/0885-064X(90)90025-9
[15] Dudley, R. M. (1973). Sample functions of the Gaussian process. Ann. Probab. 1 66-103. · Zbl 0261.60033 · doi:10.1214/aop/1176997026
[16] Efroïmovich, S. and Low, M. (1996). On optimal adaptive estimation of a quadratic functional. Ann. Statist. 24 1106-1125. · Zbl 0865.62024 · doi:10.1214/aos/1032526959
[17] Gayraud, G. and Tribouley, K. (1999). Wavelet methods to estimate an integrated quadratic functional: adaptivity and asymptotic law. Statist. Probab. Lett. 44 109-122. · Zbl 0947.62029 · doi:10.1016/S0167-7152(98)00296-X
[18] Laurent, B. (1996). Efficient estimation of integral functionals of a density. Ann. Statist. 24 659-681. · Zbl 0859.62038 · doi:10.1214/aos/1032894458
[19] Lepskii, O. V. (1990). On a problem of adaptive estimation in Gaussian white noise. Theory Probab. Appl. 35 454-466. · Zbl 0725.62075
[20] Lepskii, O. V. (1992). On problems of adaptive estimation in Gaussian white noise. Adv. Soviet Math. 12 87-106. · Zbl 0783.62061
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.