General Uncertainty Estimation with Delta Variances

TL;DR

This paper introduces Delta Variances, an efficient and versatile method for epistemic uncertainty estimation in neural networks, applicable to complex functions and requiring minimal modifications.

Contribution

It presents Delta Variances as a novel, computationally efficient approach for uncertainty quantification that unifies and extends existing techniques.

Findings

Delta Variances achieve competitive uncertainty estimates with a single gradient computation.

The method requires no changes to neural network architecture or training.

A natural extension of Delta Variances shows empirical benefits.

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.