Information Aware Max-Norm Dirichlet Networks for Predictive Uncertainty Estimation

TL;DR

This paper introduces a novel neural network method that explicitly models predictive uncertainty using Dirichlet priors, significantly improving confidence estimation and out-of-distribution detection over existing approaches.

Contribution

The proposed Information Aware Dirichlet networks learn explicit Dirichlet priors on predictions, enhancing uncertainty estimation and out-of-distribution detection compared to Bayesian neural networks.

Findings

Outperforms state-of-the-art methods in uncertainty estimation

Effectively detects out-of-distribution samples

Improves adversarial example detection

Abstract

Precise estimation of uncertainty in predictions for AI systems is a critical factor in ensuring trust and safety. Deep neural networks trained with a conventional method are prone to over-confident predictions. In contrast to Bayesian neural networks that learn approximate distributions on weights to infer prediction confidence, we propose a novel method, Information Aware Dirichlet networks, that learn an explicit Dirichlet prior distribution on predictive distributions by minimizing a bound on the expected max norm of the prediction error and penalizing information associated with incorrect outcomes. Properties of the new cost function are derived to indicate how improved uncertainty estimation is achieved. Experiments using real datasets show that our technique outperforms, by a large margin, state-of-the-art neural networks for estimating within-distribution and out-of-distribution uncertainty, and detecting adversarial examples.