Browse by author
Lookup NU author(s): Takuo Matsubara, Professor Chris Oates
This work is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0).
© 2021 Takuo Matsubara, Chris J. Oates and Francois-Xavier Briol. Bayesian neural networks attempt to combine the strong predictive performance of neural networks with formal quantification of uncertainty associated with the predictive output in the Bayesian framework. However, it remains unclear how to endow the parameters of the network with a prior distribution that is meaningful when lifted into the output space of the network. A possible solution is proposed that enables the user to posit an appropriate Gaussian process covariance function for the task at hand. Our approach constructs a prior distribution for the parameters of the network, called a ridgelet prior, that approximates the posited Gaussian process in the output space of the network. In contrast to existing work on the connection between neural networks and Gaussian processes, our analysis is non-asymptotic, with finite sample-size error bounds provided. This establishes the universality property that a Bayesian neural network can approximate any Gaussian process whose covariance function is sufficiently regular. Our experimental assessment is limited to a proof-of-concept, where we demonstrate that the ridgelet prior can out-perform an unstructured prior on regression problems for which a suitable Gaussian process prior can be provided.
Author(s): Matsubara T, Oates CJ, Briol F-X
Publication type: Article
Publication status: Published
Journal: Journal of Machine Learning Research
Year: 2021
Volume: 22
Pages: 1-57
Online publication date: 30/06/2021
Acceptance date: 31/05/2021
Date deposited: 26/08/2021
ISSN (print): 1532-4435
ISSN (electronic): 1533-7928
Publisher: Microtome Publishing
URL: https://www.jmlr.org/papers/volume22/20-1300/20-1300.pdf
