Abstract
We consider the problem of Bayesian semiparametric inference and aim to obtain an upper bound on the error of Gaussian approximation of the posterior distribution for the target parameter. This type of result can be seen as a nonasymptotic version of semiparametric Bernstein–von Mises (BvM). The provided bound is explicit in the dimension of the target parameter and in the dimension of sieve approximation of the full parameter. As a result, we can introduce the so-called critical dimension \(\,p_n\,\) of the sieve approximation, the maximal dimension for which the BvM result remains valid. In various particular statistical models, we show the necessity of the condition “\(\,p_n^2 q / n\,\) is small”, where \(\,q\,\) is the dimension of the target parameter and \(\,n\,\) is the sample size, for the BvM result to be valid under the general assumptions on the model.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Andresen, A., Spokoiny, V.: Critical dimension in profile semiparametric estimation. Electron. J. Statist. 8(2), 3077–3125 (2014). https://doi.org/10.1214/14-EJS982
Barron, A., Schervish, M.J., Wasserman, L.: The consistency of posterior distributions in nonparametric problems. Ann. Stat. 27, 536–561 (1996). https://doi.org/10.1214/aos/1018031206
Belloni, A., Chernozhukov, V.: Posterior inference in curved exponential families under increasing dimensions. Econ. J. 17(2), S75–S100 (2014). https://doi.org/10.1111/ectj.12027
Bernstein, S.: Lecture Notes on Probability Theory. Kharkiv University (1917)
Bickel, P.J., Kleijn, B.J.K.: The semiparametric Bernstein-von Mises theorem. Ann. Stat. 40(1), 206–237 (2012). https://doi.org/10.1214/11-AOS921
Bochkina, N.: Consistency of the posterior distribution in generalized linear inverse problems. Inverse Probl. 29(9), 095010 (2013). https://doi.org/10.1088/0266-5611/29/9/095010
Boucheron, S., Gassiat, E.: A Bernstein-von Mises theorem for discrete probability distributions. Electron. J. Stat. 3, 114–148 (2009). https://doi.org/10.1214/08-EJS262
Boucheron, S., Massart, P.: A high-dimensional Wilks phenomenon. Probab. Theory Relat. Fields. 150, 405–433 (2011). https://doi.org/10.1007/s00440-010-0278-7
Buhlmann, P., van de Geer, S.: Statistics for High-Dimensional Data: Methods, Theory and Applications, 1st edn. Springer Publishing Company, Incorporated (2011). https://doi.org/10.1007/978-3-642-20192-9
Burnashev, M.V.: Investigation of second order properties of statistical estimators in a scheme of independent observations. Izv. Akad. Nauk USSR Ser. Mat. 45(3), 509–539 (1981)
Castillo, I.: A semiparametric Bernstein—von Mises theorem for Gaussian process priors. Probab. Theory Relat. Fields. 152, 53–99 (2012). https://doi.org/10.1007/s00440-010-0316-5
Castillo, I., Nickl, R.: Nonparametric Bernstein-von Mises theorems in Gaussian white noise. Ann. Stat. 41(4), 1999–2028 (2013). https://doi.org/10.1214/13-AOS1133
Castillo, I., Rousseau, J.: A general bernstein–von mises theorem in semiparametric models (2013). arXiv:1305.4482 [math.ST]
Cheng, G., Kosorok, M.R.: General frequentist properties of the posterior profile distribution. Ann. Stat. 36(4), 1819–1853 (2008). https://doi.org/10.1214/07-AOS536
Chentsov, N.N.: A bound for an unknown distribution density in terms of the observations. Dokl. Akad. Nauk USSR 147(1), 45–48 (1962)
Chernozhukov, V., Hong, H.: An mcmc approach to classical estimation. J. Econ. 115(2), 293–346 (2003). https://doi.org/10.1016/S0304-4076(03)00100-3
Cox, D.D.: An analysis of Bayesian inference for nonparametric regression. Ann. Stat. 21(2), 903–923 (1993). https://doi.org/10.1214/aos/1176349157
Ermakov, M.: On semiparametric statistical inferences in the moderate deviation zone. J. Math. Sci. 152(6), 869–874 (2008). https://doi.org/10.1007/s10958-008-9104-5
Freedman, D.: On the Bernstein-von Mises theorem with infinite-dimensional parameters. Ann. Stat. 27(4), 1119–1140 (1999). https://doi.org/10.1214/aos/1017938917
Ghosal, S.: Asymptotic normality of posterior distributions in high-dimensional linear models. Bernoulli 5(2), 315–331 (1999). https://doi.org/10.2307/3318438
Gusev S.I.: Asymptotic expansions associated with some statistical estimators in the smooth case. i. expansions of random variables. Theory Probab. Appl. 20(3), 488–514 (1975)
Gusev S.I.: Asymptotic expansions associated with some statistical estimators in the smooth case. ii. expansions of moments and distributions. Theory Probab. Appl. 21(1), 16–33 (1976)
Ibragimov, I., Khas’minskij, R.: Statistical Estimation. Springer-Verlag, Asymptotic theory. New York - Heidelberg -Berlin (1981)
Kim, Y.: The Bernstein—von Mises theorem for the proportional hazard model. Ann. Stat. 34(4), 1678–1700 (2006). https://doi.org/10.1214/009053606000000533
Kim, Y., Lee, J.: A Bernstein—von Mises theorem in the nonparametric right-censoring model. Ann. Stat. 32(4), 1492–1512 (2004). https://doi.org/10.1214/009053604000000526
Kleijn, B.J.K., van der Vaart, A.W.: Misspecification in infinite-dimensional Bayesian statistics. Ann. Stat. 34(2), 837–877 (2006). https://doi.org/10.1214/009053606000000029
Kleijn, B.J.K., van der Vaart, A.W.: The Bernstein-von-Mises theorem under misspecification. Electron. J. Stat. 6, 354–381 (2012). https://doi.org/10.1214/12-EJS675
Kosorok, M.R.: Springer series in statistics. Introduction to Empirical Processes and Semiparametric Inference (2008). https://doi.org/10.1007/978-0-387-74978-5
Laurent, B., Massart, P.: Adaptive estimation of a quadratic functional by model selection. Ann. Stat. 28(5), 1302–1338 (2000). https://doi.org/10.1214/aos/1015957395
Le Cam, L., Yang, G.L.: Asymptotics in Statistics: Some Basic Concepts. Springer in Statistics (1990). https://doi.org/10.1007/978-1-4612-1166-2
Leahu, H.: On the bernstein-von mises phenomenon in the gaussian white noise model. Electron. J. Stat. 5, 373–404 (2011). https://doi.org/10.1214/11-EJS611
Mises, R.: Wahrscheinlichkeitsrechnung und ihre Anwendung in der Statistik und theoretischen Physik. Mary S, Rosenberg (1931)
Panov, M.: Nonasymptotic approach to bayesian semiparametric inference. Dokl. Math. 93(2), 155–158 (2016). https://doi.org/10.1134/S1064562416020101
Panov, M., Spokoiny, V.: Finite sample Bernstein—von Mises theorem for semiparametric problems. Bayesian Anal. 10(3), 665–710 (2015). https://doi.org/10.1214/14-BA926
Rivoirard, V., Rousseau, J.: Bernstein—von Mises theorem for linear functionals of the density. Ann. Stat. 40(3), 1489–1523 (2012). https://doi.org/10.1214/12-AOS1004
Rivoirard, V., Rousseau, J.: Posterior concentration rates for infinite dimensional exponential families. Bayesian Anal. 7(2), 311–334 (2012). https://doi.org/10.1214/12-BA710
Schwartz, L.: On bayes procedures. Probab. Theory Relat. Fields. 4(1), 10–26 (1965)
Shen, X.: Asymptotic normality of semiparametric and nonparametric posterior distributions. J. Am. Stat. Assoc. 97(457), 222–235 (2002). https://doi.org/10.1198/016214502753479365
Spokoiny, V.: Parametric estimation. Finite sample theory. Ann. Stat. 40(6), 2877–2909 (2012). https://doi.org/10.1214/12-AOS1054
Spokoiny, V.: Bernstein—von Mises Theorem for growing parameter dimension (2013). Manuscript. arXiv:1302.3430
Spokoiny, V.: Penalized maximum likelihood estimation and effective dimension. AIHP (2015). https://doi.org/10.1214/15-AIHP720
Spokoiny, V., Panov, M.: Accuracy of gaussian approximation in nonparametric bernstein–von mises theorem (2019). arXiv preprint arXiv:1910.06028
Spokoiny, V., Zhilova, M.: Sharp deviation bounds for quadratic forms. Math. Methods Stat. 22(2), 100–113 (2013). https://doi.org/10.3103/S1066530713020026
van der Vaart, A.W.: Asymptotic Statistics. Cambridge University Press (2000). https://doi.org/10.1017/CBO9780511802256
Acknowledgements
The research was supported by the Russian Science Foundation grant 20-71-10135.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Panov, M. (2023). On Accuracy of Gaussian Approximation in Bayesian Semiparametric Problems. In: Belomestny, D., Butucea, C., Mammen, E., Moulines, E., Reiß, M., Ulyanov, V.V. (eds) Foundations of Modern Statistics. FMS 2019. Springer Proceedings in Mathematics & Statistics, vol 425. Springer, Cham. https://doi.org/10.1007/978-3-031-30114-8_11
Download citation
DOI: https://doi.org/10.1007/978-3-031-30114-8_11
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-30113-1
Online ISBN: 978-3-031-30114-8
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)