Multivariate Bayesian Last Layer for Regression: Uncertainty Quantification and Disentanglement

arXiv:2405.01761v1 Announce Type: cross
Abstract: We present new Bayesian Last Layer models in the setting of multivariate regression under heteroscedastic noise, and propose an optimization algorithm for parameter learning. Bayesian Last Layer combines Bayesian modelling of the predictive distribution with neural networks for parameterization of the prior, and 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 transfer a canonically …

abstract algorithm arxiv bayesian cs.lg distribution layer modelling multivariate networks neural networks noise optimization predictive prior quantification regression stat.ml type uncertainty