Bayesian inference with sources of uncertainty: from confidence modelling to sparse estimation
Methodology
2026-05-06 v1 Machine Learning
Abstract
We introduce a general framework that extends Bayesian inference by allowing the researcher to explicitly encode confidence in each source of uncertainty within the model. This mechanism provides a new handle for model design and regularisation control. Building on this framework, we develop a general approach for inducing sparsity in statistical models and illustrate its use in linear and logistic regression, as well as in Bayesian neural networks.
Cite
@article{arxiv.2605.03134,
title = {Bayesian inference with sources of uncertainty: from confidence modelling to sparse estimation},
author = {Rafael Mouallem Rosa and Julyan Arbel and Hien Duy Nguyen},
journal= {arXiv preprint arXiv:2605.03134},
year = {2026}
}
Comments
36 pages, 8 figures, 6 tables; includes Supplementary Material