User-friendly introduction to PAC-Bayes bounds
P Alquier - Foundations and Trends® in Machine Learning, 2024 - nowpublishers.com
Aggregated predictors are obtained by making a set of basic predictors vote according to
some weights, that is, to some probability distribution. Randomized predictors are obtained …
some weights, that is, to some probability distribution. Randomized predictors are obtained …
A primer on PAC-Bayesian learning
B Guedj - arXiv preprint arXiv:1901.05353, 2019 - arxiv.org
Generalised Bayesian learning algorithms are increasingly popular in machine learning,
due to their PAC generalisation properties and flexibility. The present paper aims at …
due to their PAC generalisation properties and flexibility. The present paper aims at …
A unified recipe for deriving (time-uniform) PAC-Bayes bounds
We present a unified framework for deriving PAC-Bayesian generalization bounds. Unlike
most previous literature on this topic, our bounds are anytime-valid (ie, time-uniform) …
most previous literature on this topic, our bounds are anytime-valid (ie, time-uniform) …
On the properties of variational approximations of Gibbs posteriors
The PAC-Bayesian approach is a powerful set of techniques to derive nonasymptotic risk
bounds for random estimators. The corresponding optimal distribution of estimators, usually …
bounds for random estimators. The corresponding optimal distribution of estimators, usually …
Online pac-bayes learning
M Haddouche, B Guedj - Advances in Neural Information …, 2022 - proceedings.neurips.cc
Most PAC-Bayesian bounds hold in the batch learning setting where data is collected at
once, prior to inference or prediction. This somewhat departs from many contemporary …
once, prior to inference or prediction. This somewhat departs from many contemporary …
PAC-Bayes generalisation bounds for heavy-tailed losses through supermartingales
M Haddouche, B Guedj - arXiv preprint arXiv:2210.00928, 2022 - arxiv.org
While PAC-Bayes is now an established learning framework for light-tailed losses (\emph
{eg}, subgaussian or subexponential), its extension to the case of heavy-tailed losses …
{eg}, subgaussian or subexponential), its extension to the case of heavy-tailed losses …
Bayesian fractional posteriors
Bayesian fractional posteriors Page 1 The Annals of Statistics 2019, Vol. 47, No. 1, 39–66
https://doi.org/10.1214/18-AOS1712 © Institute of Mathematical Statistics, 2019 BAYESIAN …
https://doi.org/10.1214/18-AOS1712 © Institute of Mathematical Statistics, 2019 BAYESIAN …
Generalization bounds: Perspectives from information theory and PAC-Bayes
A fundamental question in theoretical machine learning is generalization. Over the past
decades, the PAC-Bayesian approach has been established as a flexible framework to …
decades, the PAC-Bayesian approach has been established as a flexible framework to …
On PAC-Bayesian reconstruction guarantees for VAEs
BE Chérief-Abdellatif, Y Shi… - International …, 2022 - proceedings.mlr.press
Despite its wide use and empirical successes, the theoretical understanding and study of the
behaviour and performance of the variational autoencoder (VAE) have only emerged in the …
behaviour and performance of the variational autoencoder (VAE) have only emerged in the …
Simpler PAC-Bayesian bounds for hostile data
PAC-Bayesian learning bounds are of the utmost interest to the learning community. Their
role is to connect the generalization ability of an aggregation distribution ρ ρ to its empirical …
role is to connect the generalization ability of an aggregation distribution ρ ρ to its empirical …