Timezone: »

Learning Stochastic Majority Votes by Minimizing a PAC-Bayes Generalization Bound
Valentina Zantedeschi · Paul Viallard · Emilie Morvant · Rémi Emonet · Amaury Habrard · Pascal Germain · Benjamin Guedj

Tue Dec 07 08:30 AM -- 10:00 AM (PST) @ None #None

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.

Author Information

Valentina Zantedeschi (INRIA & UCL)
Paul Viallard (University of Saint-Etienne, Lab Hubert Curien)
Emilie Morvant (LaHC, University of Saint-Etienne)
Rémi Emonet (Hubert Curien Lab.)
Amaury Habrard (University of Saint-Etienne, Lab. H Curien, France)
Pascal Germain (Université Laval)
Benjamin Guedj (Inria & University College London)

Benjamin Guedj is a tenured research scientist at Inria since 2014, affiliated to the Lille - Nord Europe research centre in France. He is also affiliated with the mathematics department of the University of Lille. Since 2018, he is a Principal Research Fellow at the Centre for Artificial Intelligence and Department of Computer Science at University College London. He is also a visiting researcher at The Alan Turing Institute. Since 2020, he is the founder and scientific director of The Inria London Programme, a strategic partnership between Inria and UCL as part of a France-UK scientific initiative. He obtained his Ph.D. in mathematics in 2013 from UPMC (Université Pierre & Marie Curie, France) under the supervision of Gérard Biau and Éric Moulines. Prior to that, he was a research assistant at DTU Compute (Denmark). His main line of research is in statistical machine learning, both from theoretical and algorithmic perspectives. He is primarily interested in the design, analysis and implementation of statistical machine learning methods for high dimensional problems, mainly using the PAC-Bayesian theory.

More from the Same Authors