Evaluating Approximate Inference in Bayesian Deep Learning

Andrew Gordon Wilson, Pavel Izmailov, Matthew D. Hoffman, Yarin Gal, Yingzhen Li, Melanie F. Pradier, Sharad Vikram, Andrew Y. K. Foong, Sanae Lotfi, Sebastian Farquhar

Abstract: Uncertainty representation is crucial to the safe and reliable deployment of deep learning. Bayesian methods provide a natural mechanism to represent epistemic uncertainty, leading to improved gene...