Posterior Temperature Optimization in Variational Inference for Inverse Problems
Image and Video Processing
2021-12-02 v3 Machine Learning
Abstract
Bayesian methods feature useful properties for solving inverse problems, such as tomographic reconstruction. The prior distribution introduces regularization, which helps solving the ill-posed problem and reduces overfitting. In practice, this often results in a suboptimal posterior temperature and the full potential of the Bayesian approach is not realized. In this paper, we optimize both the parameters of the prior distribution and the posterior temperature using Bayesian optimization. Well-tempered posteriors lead to better predictive performance and improved uncertainty calibration, which we demonstrate for the task of sparse-view CT reconstruction.
Cite
@article{arxiv.2106.07533,
title = {Posterior Temperature Optimization in Variational Inference for Inverse Problems},
author = {Max-Heinrich Laves and Malte Tölle and Alexander Schlaefer and Sandy Engelhardt},
journal= {arXiv preprint arXiv:2106.07533},
year = {2021}
}
Comments
Accepted at Bayesian Deep Learning workshop, NeurIPS 2021