Uncertainty Estimation based on Geometric Separation

TL;DR

This paper introduces a geometric-based method for uncertainty estimation in machine learning, leveraging the distance from training data to improve accuracy, calibration, and real-time applicability in critical applications.

Contribution

The paper presents a novel geometric approach for uncertainty estimation that enhances accuracy and efficiency, suitable for large-scale, real-time applications.

Findings

More accurate uncertainty estimates than recent methods

Effective calibration using post-hoc techniques

Suitable for near real-time, large dataset scenarios

Abstract

In machine learning, accurately predicting the probability that a specific input is correct is crucial for risk management. This process, known as uncertainty (or confidence) estimation, is particularly important in mission-critical applications such as autonomous driving. In this work, we put forward a novel geometric-based approach for improving uncertainty estimations in machine learning models. Our approach involves using the geometric distance of the current input from existing training inputs as a signal for estimating uncertainty, and then calibrating this signal using standard post-hoc techniques. We demonstrate that our method leads to more accurate uncertainty estimations than recently proposed approaches through extensive evaluation on a variety of datasets and models. Additionally, we optimize our approach so that it can be implemented on large datasets in near real-time applications, making it suitable for time-sensitive scenarios.