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

Zantedeschi, Valentina; Viallard, Paul; Morvant, Emilie; Emonet, Rémi; Habrard, Amaury; Germain, Pascal; Guedj, Benjamin

Type de document :

Autre communication scientifique (congrès sans actes - poster - séminaire...): Communication dans un congrès avec actes

Titre :

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

Auteur(s) :

Zantedeschi, Valentina [Auteur]
MOdel for Data Analysis and Learning [MODAL]
The Inria London Programme [Inria-London]
Viallard, Paul [Auteur]
Laboratoire Hubert Curien [LabHC]
Morvant, Emilie [Auteur]
Laboratoire Hubert Curien [LabHC]
Emonet, Rémi [Auteur]
Laboratoire Hubert Curien [LabHC]
Habrard, Amaury [Auteur]
Laboratoire Hubert Curien [LabHC]
Germain, Pascal [Auteur]
Université Laval [Québec] [ULaval]
Guedj, Benjamin [Auteur]

Department of Computer science [University College of London] [UCL-CS]
Inria-CWI [Inria-CWI]
MOdel for Data Analysis and Learning [MODAL]

Titre de la manifestation scientifique :

CAp 2022

Ville :

Vannes

Pays :

France

Date de début de la manifestation scientifique :

2022-07-05

Discipline(s) HAL :

Statistiques [stat]/Machine Learning [stat.ML]

Résumé en anglais : [en]

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 ...
Lire la suite >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.Lire moins >

Langue :

Anglais

Comité de lecture :

Oui

Audience :

Nationale

Vulgarisation :

Non

Projet ANR :

Une Perspective PAC-Bayésienne de l'Apprentissage de Représentations
Apprentissage PAC-bayésien agnostique

Collections :

Laboratoire Paul Painlevé - UMR 8524

Source :

Harvested from HAL

Learning Stochastic Majority Votes by ... BibTeX CSV Excel RIS

Learning Stochastic Majority Votes by ...

BibTeX

CSV

Excel

RIS