Saved in:
Bibliographic Details
Main Authors: Wang, Han, Kawasaki, Eiji, Damblin, Guillaume, Daniel, Geoffrey
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2405.01761
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912862238146560
author Wang, Han
Kawasaki, Eiji
Damblin, Guillaume
Daniel, Geoffrey
author_facet Wang, Han
Kawasaki, Eiji
Damblin, Guillaume
Daniel, Geoffrey
contents We present new Bayesian Last Layer neural network models in the setting of multivariate regression under heteroscedastic noise, and propose EM algorithms for parameter learning. Bayesian modeling of a neural network's final layer has the attractive property of uncertainty quantification with a single forward pass. The proposed framework is capable of disentangling the aleatoric and epistemic uncertainty, and can be used to enhance a canonically trained deep neural network with uncertainty-aware capabilities.
format Preprint
id arxiv_https___arxiv_org_abs_2405_01761
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Multivariate Bayesian Last Layer for Regression with Uncertainty Quantification and Decomposition
Wang, Han
Kawasaki, Eiji
Damblin, Guillaume
Daniel, Geoffrey
Machine Learning
We present new Bayesian Last Layer neural network models in the setting of multivariate regression under heteroscedastic noise, and propose EM algorithms for parameter learning. Bayesian modeling of a neural network's final layer has the attractive property of uncertainty quantification with a single forward pass. The proposed framework is capable of disentangling the aleatoric and epistemic uncertainty, and can be used to enhance a canonically trained deep neural network with uncertainty-aware capabilities.
title Multivariate Bayesian Last Layer for Regression with Uncertainty Quantification and Decomposition
topic Machine Learning
url https://arxiv.org/abs/2405.01761