Schmitt, S; Shawe-Taylor, J; van Hasselt, H; (2025) General Uncertainty Estimation with Delta Variances. In: Proceedings of the AAAI Conference on Artificial Intelligence. (pp. pp. 20318-20328). Association for the Advancement of Artificial Intelligence (AAAI)

Preview

Text 34238-Article Text-38306-1-2-20250410.pdf - Published Version Download (299kB) | Preview

Abstract

Decision makers may suffer from uncertainty induced by limited data. This may be mitigated by accounting for epistemic uncertainty, which is however challenging to estimate efficiently for large neural networks. To this extent we investigate Delta Variances, a family of algorithms for epistemic uncertainty quantification, that is computationally efficient and convenient to implement. It can be applied to neural networks and more general functions composed of neural networks. As an example we consider a weather simulator with a neural-network-based step function inside - here Delta Variances empirically obtain competitive results at the cost of a single gradient computation. The approach is convenient as it requires no changes to the neural network architecture or training procedure. We discuss multiple ways to derive Delta Variances theoretically noting that special cases recover popular techniques and present a unified perspective on multiple related methods. Finally we observe that this general perspective gives rise to a natural extension and empirically show its benefit.

Download activity - last month

Export as XMLCSVJSON

Download activity - last 12 months

Export as XMLCSVJSON

Downloads by country - last 12 months

Export as CSVXMLJSON

Archive Staff Only

View Item