We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.LG

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Machine Learning

Title: Improving predictions of Bayesian neural nets via local linearization

Abstract: The generalized Gauss-Newton (GGN) approximation is often used to make practical Bayesian deep learning approaches scalable by replacing a second order derivative with a product of first order derivatives. In this paper we argue that the GGN approximation should be understood as a local linearization of the underlying Bayesian neural network (BNN), which turns the BNN into a generalized linear model (GLM). Because we use this linearized model for posterior inference, we should also predict using this modified model instead of the original one. We refer to this modified predictive as "GLM predictive" and show that it effectively resolves common underfitting problems of the Laplace approximation. It extends previous results in this vein to general likelihoods and has an equivalent Gaussian process formulation, which enables alternative inference schemes for BNNs in function space. We demonstrate the effectiveness of our approach on several standard classification datasets as well as on out-of-distribution detection. We provide an implementation at this https URL
Comments: AISTATS 2021
Subjects: Machine Learning (stat.ML); Machine Learning (cs.LG)
Cite as: arXiv:2008.08400 [stat.ML]
  (or arXiv:2008.08400v3 [stat.ML] for this version)

Submission history

From: Alexander Immer [view email]
[v1] Wed, 19 Aug 2020 12:35:55 GMT (4828kb,D)
[v2] Thu, 22 Oct 2020 14:45:01 GMT (4629kb,D)
[v3] Thu, 25 Feb 2021 17:59:47 GMT (5493kb,D)

Link back to: arXiv, form interface, contact.