Scalable Utility-Aware Multiclass Calibration

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Hegazy, Mahmoud, Jordan, Michael I., Dieuleveut, Aymeric
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909876196737024
author Hegazy, Mahmoud
Jordan, Michael I.
Dieuleveut, Aymeric
author_facet Hegazy, Mahmoud
Jordan, Michael I.
Dieuleveut, Aymeric
contents Ensuring that classifiers are well-calibrated, i.e., their predictions align with observed frequencies, is a minimal and fundamental requirement for classifiers to be viewed as trustworthy. Existing methods for assessing multiclass calibration often focus on specific aspects associated with prediction (e.g., top-class confidence, class-wise calibration) or utilize computationally challenging variational formulations. In this work, we study scalable \emph{evaluation} of multiclass calibration. To this end, we propose utility calibration, a general framework that measures the calibration error relative to a specific utility function that encapsulates the goals or decision criteria relevant to the end user. We demonstrate how this framework can unify and re-interpret several existing calibration metrics, particularly allowing for more robust versions of the top-class and class-wise calibration metrics, and, going beyond such binarized approaches, toward assessing calibration for richer classes of downstream utilities.
format Preprint
id arxiv_https___arxiv_org_abs_2510_25458
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Scalable Utility-Aware Multiclass Calibration
Hegazy, Mahmoud
Jordan, Michael I.
Dieuleveut, Aymeric
Machine Learning
Artificial Intelligence
Ensuring that classifiers are well-calibrated, i.e., their predictions align with observed frequencies, is a minimal and fundamental requirement for classifiers to be viewed as trustworthy. Existing methods for assessing multiclass calibration often focus on specific aspects associated with prediction (e.g., top-class confidence, class-wise calibration) or utilize computationally challenging variational formulations. In this work, we study scalable \emph{evaluation} of multiclass calibration. To this end, we propose utility calibration, a general framework that measures the calibration error relative to a specific utility function that encapsulates the goals or decision criteria relevant to the end user. We demonstrate how this framework can unify and re-interpret several existing calibration metrics, particularly allowing for more robust versions of the top-class and class-wise calibration metrics, and, going beyond such binarized approaches, toward assessing calibration for richer classes of downstream utilities.
title Scalable Utility-Aware Multiclass Calibration
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2510.25458