On the Need to Align Intent and Implementation in Uncertainty Quantification for Machine Learning

TL;DR

This paper discusses the importance of aligning the intent and implementation of uncertainty quantification in machine learning, emphasizing standards and trustworthiness across diverse contexts and applications.

Contribution

It identifies inconsistencies in terminology and approaches in UQ, and advocates for standards that align the goals and methods of uncertainty estimation in ML.

Findings

Highlighting the need for consistent terminology and mapping in UQ

Proposing standards to align intent and implementation in UQ

Providing practical recommendations for scientific ML and SBI

Abstract

Quantifying uncertainties for machine learning (ML) models is a foundational challenge in modern data analysis. This challenge is compounded by at least two key aspects of the field: (a) inconsistent terminology surrounding uncertainty and estimation across disciplines, and (b) the varying technical requirements for establishing trustworthy uncertainties in diverse problem contexts. In this position paper, we aim to clarify the depth of these challenges by identifying these inconsistencies and articulating how different contexts impose distinct epistemic demands. We examine the current landscape of estimation targets (e.g., prediction, inference, simulation-based inference), uncertainty constructs (e.g., frequentist, Bayesian, fiducial), and the approaches used to map between them. Drawing on the literature, we highlight and explain examples of problematic mappings. To help address these issues, we advocate for standards that promote alignment between the \textit{intent} and \textit{implementation} of uncertainty quantification (UQ) approaches. We discuss several axes of trustworthiness that are necessary (if not sufficient) for reliable UQ in ML models, and show how these axes can inform the design and evaluation of uncertainty-aware ML systems. Our practical recommendations focus on scientific ML, offering illustrative cases and use scenarios, particularly in the context of simulation-based inference (SBI).