Learning Stochastic Majority Votes by Minimizing a PAC-Bayes Generalization Bound - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2022

Learning Stochastic Majority Votes by Minimizing a PAC-Bayes Generalization Bound

Abstract

We investigate a stochastic counterpart of majority votes over finite ensembles of classifiers, and study its generalization properties. While our approach holds for arbitrary distributions, we instantiate it with Dirichlet distributions: this allows for a closed-form and differentiable expression for the expected risk, which then turns the generalization bound into a tractable training objective. The resulting stochastic majority vote learning algorithm achieves state-of-the-art accuracy and benefits from (non-vacuous) tight generalization bounds, in a series of numerical experiments when compared to competing algorithms which also minimize PAC-Bayes objectives - both with uninformed (data-independent) and informed (data-dependent) priors.
No file

Dates and versions

hal-03703804 , version 1 (24-06-2022)

Identifiers

  • HAL Id : hal-03703804 , version 1

Cite

Valentina Zantedeschi, Paul Viallard, Emilie Morvant, Rémi Emonet, Amaury Habrard, et al.. Learning Stochastic Majority Votes by Minimizing a PAC-Bayes Generalization Bound. CAp 2022, Jul 2022, Vannes, France. ⟨hal-03703804⟩
36 View
0 Download

Share

Gmail Facebook X LinkedIn More