PAC-Bayesian Bounds on Constrained f-Entropic Risk Measures

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Atbir, Hind, Cherfaoui, Farah, Metzler, Guillaume, Morvant, Emilie, Viallard, Paul
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910111140675584
author Atbir, Hind
Cherfaoui, Farah
Metzler, Guillaume
Morvant, Emilie
Viallard, Paul
author_facet Atbir, Hind
Cherfaoui, Farah
Metzler, Guillaume
Morvant, Emilie
Viallard, Paul
contents PAC generalization bounds on the risk, when expressed in terms of the expected loss, are often insufficient to capture imbalances between subgroups in the data. To overcome this limitation, we introduce a new family of risk measures, called constrained f-entropic risk measures, which enable finer control over distributional shifts and subgroup imbalances via f-divergences, and include the Conditional Value at Risk (CVaR), a well-known risk measure. We derive both classical and disintegrated PAC-Bayesian generalization bounds for this family of risks, providing the first disintegratedPAC-Bayesian guarantees beyond standard risks. Building on this theory, we design a self-bounding algorithm that minimizes our bounds directly, yielding models with guarantees at the subgroup level. Finally, we empirically demonstrate the usefulness of our approach.
format Preprint
id arxiv_https___arxiv_org_abs_2510_11169
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle PAC-Bayesian Bounds on Constrained f-Entropic Risk Measures
Atbir, Hind
Cherfaoui, Farah
Metzler, Guillaume
Morvant, Emilie
Viallard, Paul
Machine Learning
PAC generalization bounds on the risk, when expressed in terms of the expected loss, are often insufficient to capture imbalances between subgroups in the data. To overcome this limitation, we introduce a new family of risk measures, called constrained f-entropic risk measures, which enable finer control over distributional shifts and subgroup imbalances via f-divergences, and include the Conditional Value at Risk (CVaR), a well-known risk measure. We derive both classical and disintegrated PAC-Bayesian generalization bounds for this family of risks, providing the first disintegratedPAC-Bayesian guarantees beyond standard risks. Building on this theory, we design a self-bounding algorithm that minimizes our bounds directly, yielding models with guarantees at the subgroup level. Finally, we empirically demonstrate the usefulness of our approach.
title PAC-Bayesian Bounds on Constrained f-Entropic Risk Measures
topic Machine Learning
url https://arxiv.org/abs/2510.11169