Evidential Deep Learning to Quantify Classification Uncertainty

TL;DR

This paper introduces a method called Evidential Deep Learning that explicitly models uncertainty in neural network predictions using Dirichlet distributions, improving out-of-distribution detection and robustness against adversarial attacks.

Contribution

It proposes a novel approach to quantify classification uncertainty by modeling predictions as subjective opinions with Dirichlet distributions, distinct from Bayesian methods.

Findings

Enhanced detection of out-of-distribution samples

Improved robustness to adversarial perturbations

Effective uncertainty estimation in classification tasks

Abstract

Deterministic neural nets have been shown to learn effective predictors on a wide range of machine learning problems. However, as the standard approach is to train the network to minimize a prediction loss, the resultant model remains ignorant to its prediction confidence. Orthogonally to Bayesian neural nets that indirectly infer prediction uncertainty through weight uncertainties, we propose explicit modeling of the same using the theory of subjective logic. By placing a Dirichlet distribution on the class probabilities, we treat predictions of a neural net as subjective opinions and learn the function that collects the evidence leading to these opinions by a deterministic neural net from data. The resultant predictor for a multi-class classification problem is another Dirichlet distribution whose parameters are set by the continuous output of a neural net. We provide a preliminary analysis on how the peculiarities of our new loss function drive improved uncertainty estimation. We observe that our method achieves unprecedented success on detection of out-of-distribution queries and endurance against adversarial perturbations.