Volltext-Downloads (blau) und Frontdoor-Views (grau)

Bernstein flows for flexible posteriors in variational Bayes

  • Black-box variational inference (BBVI) is a technique to approximate the posterior of Bayesian models by optimization. Similar to MCMC, the user only needs to specify the model; then, the inference procedure is done automatically. In contrast to MCMC, BBVI scales to many observations, is faster for some applications, and can take advantage of highly optimized deep learning frameworks since it can be formulated as a minimization task. In the case of complex posteriors, however, other state-of-the-art BBVI approaches often yield unsatisfactory posterior approximations. This paper presents Bernstein flow variational inference (BF-VI), a robust and easy-to-use method flexible enough to approximate complex multivariate posteriors. BF-VI combines ideas from normalizing flows and Bernstein polynomial-based transformation models. In benchmark experiments, we compare BF-VI solutions with exact posteriors, MCMC solutions, and state-of-the-art BBVI methods, including normalizing flow-based BBVI. We show for low-dimensional models that BF-VI accurately approximates the true posterior; in higher-dimensional models, BF-VI compares favorably against other BBVI methods. Further, using BF-VI, we develop a Bayesian model for the semi-structured melanoma challenge data, combining a CNN model part for image data with an interpretable model part for tabular data, and demonstrate, for the first time, the use of BBVI in semi-structured models.

Download full text files

Export metadata

Additional Services

Search Google Scholar


Author:Oliver DürrORCiDGND, Stefan Hörtling, Daniel Dold, Ivonne Kovylov, Beate SickORCiDGND
Parent Title (English):AStA Advances in Statistical Analysis
Publisher:Springer Science and Business Media LLC
Place of publication:Berlin; Heidelberg
Document Type:Article
Year of Publication:2024
Release Date:2024/04/24
Tag:Variational inference; Deep learning; Transformation models; Bayesian neural network
Page Number:20 Seiten
Open Access Publikation finanziert durch Projekt DEAL
Corresponding author: Oliver Dürr und Beate Sick
Institutes:Institut für Optische Systeme - IOS
Open Access?:Ja
Relevance:Peer reviewed nach anderen Listungen (mit Nachweis zum Peer Review Verfahren)
Licence (German):License LogoCreative Commons - CC BY - Namensnennung 4.0 International