Abstract
This is a review of asymptotic and non-asymptotic behaviour of Bayesian methods under model specification. In particular we focus on consistency, i.e. convergence of the posterior distribution to the point mass at the best parametric approximation to the true model, and conditions for it to be locally Gaussian around this point. For well specified regular models, variance of the Gaussian approximation coincides with the Fisher information, making Bayesian inference asymptotically efficient. In this review, we discuss how this is affected by model misspecification. In particular, we highlight contribution of Volodia Spokoiny to this area. We also discuss approaches to adjust Bayesian inference to make it asymptotically efficient under model misspecification.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Baraud, Y., Birgé, L.: Robust Bayes-like estimation: Rho-Bayes estimation. Ann. Stat. 48(6), 3699–3720, 12 2020
Besag, J.: On the statistical analysis of dirty pictures (with discussion). J. Roy. Statist. Soc. B 48, 259–302 (1986)
Bhattacharya, A., Pati, D., Yang, Y.: Bayesian fractional posteriors. Ann. Statist. 47(1), 39–66, 02 2019
Bhattacharya, I., Ghosal, S.: Bayesian inference on multivariate medians and quantiles. Statistica Sinica (2019)
Bhattacharya, I., Martin, R.: Gibbs posterior inference on multivariate quantiles. J. Stat. Plann. Infer. 218, 106–121 (2022)
Bissiri, P.G., Holmes, C.C., Walker, S.G.: A general framework for updating belief distributions. J. R. Statist. Soc.: Ser. B (Statistical Methodology) (2016)
Bochkina, N.A., Green, P.J.: The Bernstein–von Mises theorem and nonregular models. Ann. Statist. 42(5), 1850–1878, 10 2014
Breiman, L.: Bagging predictors. Mach. Learn. 24, 123–140 (1996)
Chernozhukov, V., Hong, H.: Likelihood estimation and inference in a class of nonregular econometric models. Econometrica 72, 1445–1480 (2004)
Chib, S., Shin, M., Simoni, A.: Bayesian estimation and comparison of moment condition models. J. Am. Statist. Assoc. 113(524), 1656–1668 (2018)
Dalalyan, A., Tsybakov, A.B.: Aggregation by exponential weighting, sharp PAC-Bayesian bounds and sparsity. Mach. Learn. 72, 39–61 (2008)
de Heide, R., Kirichenko, A., Mehta, N., Grünwald, P.: Safe-Bayesian generalized linear regression. In: Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics, volume 108 of Proceedings of Machine Learning Research, pp. 2623 –2633. PMLR (2020)
Diong, M.L., Chaumette, E., Vincent, F.: On the efficiency of maximum-likelihood estimators of misspecified models. In: 25th European Signal Processing Conference (EUSIPCO) (2017)
Fong, E., Holmes, C., Walker, S.G.: Martingale posterior distributions. J. R. Statist. Soc.: Ser. B (Statistical Methodology) (2023).
Grünwald, P.: The safe Bayesian. In: International Conference on Algorithmic Learning Theory, pp. 169–183. Springer (2012)
Grünwald, P., van Ommen, T.: Inconsistency of Bayesian inference for misspecified linear models, and a proposal for repairing it. Bayesian Anal. 12(4), 1069–1103, 12 2017
Grünwald, P.D., Mehta, N.A.: Fast rates for general unbounded loss functions: from ERM to generalized Bayes. J. Mach. Learn. Res. 21(56), 1–80 (2020)
Holmes, C.C., Walker, S.G.: Assigning a value to a power likelihood in a general Bayesian model. Biometrika 104(2), 497–503 (2017)
Huggins, J., Miller, J.: Reproducible model selection using bagged posteriors. Bayesian. Anal. 18(1), 79–104 (2023)
Ibragimov, I., Hasminskij, R.: Statistical Estimation: Asymptotic Theory. Springer (1981)
Kleijn, B.J.K., van der Vaart, A.W.: Misspecification in infinite-dimensional Bayesian statistics. Ann. Statist. 34(2), 837–877, 04 2006
Kleijn, B.J.K., van der Vaart, A.W.: The Bernstein–von-Mises theorem under misspecification. Electron. J. Statist. 6, 354–381 (2012)
Knoblauch, J., Jewson, J., Damoulas, T.: Generalized variational inference: Three arguments for deriving new posteriors (2021). arXiv:1904.02063
Lindsay, B.: Composite likelihood methods. Contemp. Math. 80, 221–239 (1988)
Lyddon, S.P., Holmes, C.C., Walker, S.G.: General Bayesian updating and the loss-likelihood bootstrap. Biometrika 106, 465–478 (2019)
Miller, J.W., Dunson, D.B.: Robust Bayesian inference via coarsening. J. Am. Statist. Assoc. 114(527), 1113–1125 (2019)
Müller, U.K.: Risk of Bayesian inference in misspecified models, and the sandwich covariance matrix. Econometrica 81(5), 1805–1849 (2013)
Newton, M.A., Polson, N.G., Xu, J.: Weighted Bayesian bootstrap for scalable posterior distributions. Canadian J. Statist. (2021)
Panov, M., Spokoiny, V.: Finite sample Bernstein–von Mises theorem for semiparametric problems. Bayesian Anal. 10(3), 665–710, 09 2015
Pauli, F., Racugno, W., Ventura, L.: Bayesian composite marginal likelihoods. Statistica Sinica 21, 149–164 (2012)
Ribatet, M., Cooley, D., Davison, A.C.: Bayesian inference from composite likelihoods, with an application to spatial extremes. Statistica Sinica 22, 813–845 (2012)
Rubin, D.B.: The Bayesian bootstrap. Ann. Statist. 9, 130–134 (1981)
Schennach, S.M.: Bayesian exponentially tilted empirical likelihood. Biometrika 92(1), 31–46 (2005)
Spokoiny, V.: Parametric estimation. Finite sample theory. Ann. Statist. 40(6), 2877–2909, 12 2012
Spokoiny, V.: Bayesian inference for nonlinear inverse problems (2020). arXiv:1912.12694
Spokoiny, V., Panov, M.: Accuracy of Gaussian approximation in nonparametric Bernstein–von Mises (2020). arXiv:1910.06028
Stoehr, J., Friel, N.: Calibration of conditional composite likelihood for Bayesian inference on Gibbs random fields. In: Lebanon, G., Vishwanathan, S.V.N. (eds.), Proceedings of the Eighteenth International Conference on Artificial Intelligence and Statistics, volume 38 of Proceedings of Machine Learning Research, pp. 921–929. PMLR (2015)
Syring, N., Martin, R.: Gibbs posterior concentration rates under sub-exponential type losses. Bernoulli 29(2), 1080–1108 (2023)
Van der Vaart, A.W.: Asymptotic Statistics, vol. 3. Cambridge University Press (2000)
Varin, C., Reid, N., Firth, D.: An overview of composite likelihood methods. Statistica Sinica 21, 5–42 (2011)
Waddell, P.J., Kishino, H., Ota, R.: Very fast algorithms for evaluating the stability of ml and Bayesian phylogenetic trees from sequence data. In Genome Inf. 13, 82–92 (2002)
Wang, Y., Blei, D.M.: Variational Bayes under model misspecification. In: In Advances in Neural Information Processing Systems (2019)
Wang, Y., Kucukelbir, A., Blei, D.M.: Robust probabilistic modeling with Bayesian data reweighting. In: Proceedings of the 34th International Conference on Machine Learning, volume 70 of Proceedings of Machine Learning Research, pp. 3646–3655. PMLR (2017)
White, H.: Maximum likelihood estimation of misspecified models. Econometrica 50, 1–25 (1982)
Wu, P.-S., Martin, R.: A comparison of learning rate selection methods in generalized Bayesian inference (2020). arxiv:2012.11349
Zhang, Y., Nalisnick, E.: On the inconsistency of Bayesian inference for misspecified neural networks. In: Third Symposium on Advances in Approximate Bayesian Inference (2021)
Acknowledgements
This review was in part motivated by the discussion of the author with Peter Grünwald, Pierre Jacob and Jeffrey Miller during a Research in Groups meeting sponsored by the International Centre for Mathematical Sciences in Edinburgh, UK.
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
Bochkina, N. (2023). Bernstein–von Mises Theorem and Misspecified Models: A Review. 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_10
Download citation
DOI: https://doi.org/10.1007/978-3-031-30114-8_10
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)