Large ball probabilities, Gaussian comparison and anti-concentration. (English) Zbl 1428.62187

Summary: We derive tight non-asymptotic bounds for the Kolmogorov distance between the probabilities of two Gaussian elements to hit a ball in a Hilbert space. The key property of these bounds is that they are dimension-free and depend on the nuclear (Schatten-one) norm of the difference between the covariance operators of the elements and on the norm of the mean shift. The obtained bounds significantly improve the bound based on Pinsker’s inequality via the Kullback-Leibler divergence. We also establish an anti-concentration bound for a squared norm of a non-centered Gaussian element in Hilbert space. The paper presents a number of examples motivating our results and applications of the obtained bounds to statistical inference and to high-dimensional CLT.


60B11 Probability theory on linear topological spaces
60E15 Inequalities; stochastic orderings
60F05 Central limit and other weak theorems
Full Text: DOI arXiv Euclid


[1] Ball, K. (1993). The reverse isoperimetric problem for Gaussian measure. Discrete Comput. Geom.10 411-420. · Zbl 0788.52010
[2] Barsov, S.S. and Ulyanov, V.V. (1987). Estimates of the proximity of Gaussian measures. Soviet Math. Dokl.34 462-466. · Zbl 0628.28014
[3] Belitser, E. (2017). On coverage and local radial rates of credible sets. Ann. Statist.45 1124-1151. · Zbl 1371.62044
[4] Bentkus, V. (2003). On the dependence of the Berry-Esseen bound on dimension. J. Statist. Plann. Inference113 385-402. · Zbl 1017.60023
[5] Bentkus, V. (2005). A Lyapunov-type bound in \(R^d\) . Theory Probab. Appl.49 311-323. · Zbl 1090.60019
[6] Bobkov, S.G. and Chistyakov, G.P. (2014). Bounds for the maximum of the density of the sum of independent random variables. J. Math. Sci.199 100-106. · Zbl 1320.60054
[7] Bobkov, S.G., Chistyakov, G.P. and Götze, F. (2014). Berry-Esseen bounds in the entropic central limit theorem. Probab. Theory Related Fields159 435-478. · Zbl 1307.60011
[8] Bogachev, V.I. (1998). Gaussian Measures. Mathematical Surveys and Monographs62. Providence, RI: Amer. Math. Soc.
[9] Bontemps, D. (2011). Bernstein – von Mises theorems for Gaussian regression with increasing number of regressors. Ann. Statist.39 2557-2584. · Zbl 1231.62061
[10] Castillo, I. (2012). A semiparametric Bernstein – von Mises theorem for Gaussian process priors. Probab. Theory Related Fields152 53-99. · Zbl 1232.62054
[11] Castillo, I. and Nickl, R. (2013). Nonparametric Bernstein – von Mises theorems in Gaussian white noise. Ann. Statist.41 1999-2028. · Zbl 1285.62052
[12] Christoph, G., Prokhorov, Y. and Ulyanov, V. (1996). On distribution of quadratic forms in Gaussian random variables. Theory Probab. Appl.40 250-260.
[13] Chung, K.L. (2001). A Course in Probability Theory, 3rd ed. San Diego, CA: Academic Press.
[14] Götze, F. and Zaitsev, A.Yu. (2018). New applications of Arak’s inequalities to the Littlewood-Offord problem. Eur. J. Math.4 639-663. · Zbl 1392.60041
[15] Johnson, N.L., Kotz, S. and Balakrishnan, N. (1994). Continuous Univariate Distributions. Vol. 1, 2nd ed. Wiley Series in Probability and Mathematical Statistics: Applied Probability and Statistics. New York: Wiley. A Wiley-Interscience Publication.
[16] Johnstone, I.M. (2010). High dimensional Bernstein – von Mises: Simple examples. In Borrowing Strength: Theory Powering Applications—A Festschrift for Lawrence D. Brown. Inst. Math. Stat. (IMS) Collect.6 87-98. Beachwood, OH: IMS.
[17] Ledoux, M. and Talagrand, M. (2002). Probability in Banach Spaces. Isoperimetry and Processes. Berlin: Springer. · Zbl 1226.60003
[18] Li, W.V. and Shao, Q.-M. (2001). Gaussian processes: Inequalities, small ball probabilities and applications. In Stochastic Processes: Theory and Methods. Handbook of Statist.19 533-597. Amsterdam: North-Holland. · Zbl 0987.60053
[19] Lifshits, M. (2012). Lectures on Gaussian Processes. SpringerBriefs in Mathematics. Heidelberg: Springer. · Zbl 1248.60002
[20] Markus, A.S. (1964). The eigen- and singular values of the sum and product of linear operators. Russian Math. Surveys19 91-120. · Zbl 0133.07205
[21] Naumov, A.A., Spokoiny, V.G., Tavyrikov, Yu.E. and Ulyanov, V.V. (2018). Nonasymptotic estimates for the closeness of Gaussian measures on balls. Dokl. Math.98. · Zbl 1420.60050
[22] Naumov, A.A., Spokoiny, V.G. and Ulyanov, V.V. (2018). Confidence sets for spectral projectors of covariance matrices. Dokl. Math.98. · Zbl 1409.62148
[23] Panov, M. and Spokoiny, V. (2015). Finite sample Bernstein – von Mises theorem for semiparametric problems. Bayesian Anal.10 665-710. · Zbl 1335.62057
[24] Prokhorov, Y.V. and Ulyanov, V.V. (2013). Some approximation problems in statistics and probability. In Limit Theorems in Probability, Statistics and Number Theory. Springer Proc. Math. Stat.42 235-249. Heidelberg: Springer. · Zbl 1293.62037
[25] Rudelson, M. and Vershynin, R. (2008). The Littlewood-Offord problem and invertibility of random matrices. Adv. Math.218 600-633. · Zbl 1139.15015
[26] Sazonov, V.V. (1981). Normal Approximation—Some Recent Advances. Lecture Notes in Math.879. Berlin-New York: Springer. · Zbl 0462.60006
[27] Spokoiny, V. (2017). Penalized maximum likelihood estimation and effective dimension. Ann. Inst. Henri Poincaré Probab. Stat.53 389-429. · Zbl 1421.62021
[28] Spokoiny, V. and Zhilova, M. (2015). Bootstrap confidence sets under model misspecification. Ann. Statist.43 2653-2675. · Zbl 1327.62179
[29] Tropp, J.A. (2012). User-friendly tail bounds for sums of random matrices. Found. Comput. Math.12 389-434. · Zbl 1259.60008
[30] Ul’yanov, V. (1987). Asymptotic expansions for distributions of sums of independent random variables in H. Theory Probab. Appl.31 25-39.
[31] Yurinsky, V. (1995). Sums and Gaussian Vectors. Lecture Notes in Math.1617. Berlin: Springer. · Zbl 0846.60003
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. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.