More PAC-Bayes bounds: From bounded losses, to losses with general tail
behaviors, to anytime validityJournal of machine learning research (JMLR), 2023 |
A One-Sample Decentralized Proximal Algorithm for Non-Convex Stochastic
Composite OptimizationConference on Uncertainty in Artificial Intelligence (UAI), 2023 |
Generalisation under gradient descent via deterministic PAC-BayesInternational Conference on Algorithmic Learning Theory (ALT), 2022 |
Online PAC-Bayes LearningNeural Information Processing Systems (NeurIPS), 2022 |
Dimensionality Reduction and Wasserstein Stability for Kernel RegressionJournal of machine learning research (JMLR), 2022 |
On PAC-Bayesian reconstruction guarantees for VAEsInternational Conference on Artificial Intelligence and Statistics (AISTATS), 2022 |
Risk bounds for aggregated shallow neural networks using Gaussian priorAnnual Conference Computational Learning Theory (COLT), 2021 |
Conditionally Gaussian PAC-BayesInternational Conference on Artificial Intelligence and Statistics (AISTATS), 2021 |
NTS-NOTEARS: Learning Nonparametric DBNs With Prior KnowledgeInternational Conference on Artificial Intelligence and Statistics (AISTATS), 2021 |
On Margins and Derandomisation in PAC-BayesInternational Conference on Artificial Intelligence and Statistics (AISTATS), 2021 |
Convergence guarantee for the sparse monotone single index modelElectronic Journal of Statistics (EJS), 2021 |
Learning Sparse Nonparametric DAGsInternational Conference on Artificial Intelligence and Statistics (AISTATS), 2019 |
On the properties of variational approximations of Gibbs posteriorsJournal of machine learning research (JMLR), 2015 |
PAC-Bayesian AUC classification and scoringNeural Information Processing Systems (NeurIPS), 2014 |
On Oracle Property and Asymptotic Validity of Bayesian Generalized
Method of MomentsJournal of Multivariate Analysis (JMA), 2014 |
COBRA: A Combined Regression StrategyJournal of Multivariate Analysis (J. Multivar. Anal.), 2013 |
PAC-Bayesian Estimation and Prediction in Sparse Additive ModelsElectronic Journal of Statistics (EJS), 2012 |