English

Robust PAC$^m$: Training Ensemble Models Under Misspecification and Outliers

Machine Learning 2023-04-25 v3 Information Theory math.IT Machine Learning

Abstract

Standard Bayesian learning is known to have suboptimal generalization capabilities under misspecification and in the presence of outliers. PAC-Bayes theory demonstrates that the free energy criterion minimized by Bayesian learning is a bound on the generalization error for Gibbs predictors (i.e., for single models drawn at random from the posterior) under the assumption of sampling distributions uncontaminated by outliers. This viewpoint provides a justification for the limitations of Bayesian learning when the model is misspecified, requiring ensembling, and when data is affected by outliers. In recent work, PAC-Bayes bounds -- referred to as PACm^m -- were derived to introduce free energy metrics that account for the performance of ensemble predictors, obtaining enhanced performance under misspecification. This work presents a novel robust free energy criterion that combines the generalized logarithm score function with PACm^m ensemble bounds. The proposed free energy training criterion produces predictive distributions that are able to concurrently counteract the detrimental effects of misspecification -- with respect to both likelihood and prior distribution -- and outliers.

Keywords

Cite

@article{arxiv.2203.01859,
  title  = {Robust PAC$^m$: Training Ensemble Models Under Misspecification and Outliers},
  author = {Matteo Zecchin and Sangwoo Park and Osvaldo Simeone and Marios Kountouris and David Gesbert},
  journal= {arXiv preprint arXiv:2203.01859},
  year   = {2023}
}
R2 v1 2026-06-24T10:01:09.264Z