Quantifying Calibration Error in Neural Networks Through Evidence-Based Theory

TL;DR

This paper presents a new evidence-based framework for quantifying neural network calibration, improving trustworthiness assessment by capturing confidence, disbelief, and uncertainty, demonstrated on MNIST and CIFAR-10 datasets.

Contribution

It introduces a novel subjective logic-based method for evaluating calibration error, enhancing interpretability and trustworthiness measurement of neural networks.

Findings

Improved calibration assessment on MNIST and CIFAR-10

Enhanced interpretability of model confidence and uncertainty

Potential applications in healthcare and autonomous systems

Abstract

Trustworthiness in neural networks is crucial for their deployment in critical applications, where reliability, confidence, and uncertainty play pivotal roles in decision-making. Traditional performance metrics such as accuracy and precision fail to capture these aspects, particularly in cases where models exhibit overconfidence. To address these limitations, this paper introduces a novel framework for quantifying the trustworthiness of neural networks by incorporating subjective logic into the evaluation of Expected Calibration Error (ECE). This method provides a comprehensive measure of trust, disbelief, and uncertainty by clustering predicted probabilities and fusing opinions using appropriate fusion operators. We demonstrate the effectiveness of this approach through experiments on MNIST and CIFAR-10 datasets, where post-calibration results indicate improved trustworthiness. The proposed framework offers a more interpretable and nuanced assessment of AI models, with potential applications in sensitive domains such as healthcare and autonomous systems.