{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,7]],"date-time":"2026-07-07T12:11:43Z","timestamp":1783426303402,"version":"3.54.6"},"reference-count":247,"publisher":"Emerald","issue":"2","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2024,1,22]]},"abstract":"<jats:p>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 by sampling in a set of basic predictors, according to some prescribed probability distribution.<\/jats:p>\n                  <jats:p>Thus, aggregated and randomized predictors have in common that their definition rely on a probability distribution on the set of predictors. In statistical learning theory, there is a set of tools designed to understand the generalization ability of such predictors: PAC-Bayesian or PAC-Bayes bounds.<\/jats:p>\n                  <jats:p>Since the original PAC-Bayes bounds (Shawe-Taylor and Williamson, 1997; McAllester, 1998), these tools have been considerably improved in many directions. We will for example describe a simplified version of the localization technique (Catoni, 2003; Catoni, 2007) that was missed by the community, and later rediscovered as \u201cmutual information bounds\u201d. Very recently, PAC-Bayes bounds received a considerable attention. There was workshop on PAC-Bayes at NIPS 2017, (Almost) 50 Shades of Bayesian Learning: PAC-Bayesian trends and insights, organized by B. Guedj, F. Bach and P. Germain. One of the reasons of this recent interest is the successful application of these bounds to neural networks (Dziugaite and Roy, 2017). Since then, this is a recurring topic of workshops in the major machine learning conferences.<\/jats:p>\n                  <jats:p>The objective of these notes is to provide an elementary introduction to PAC-Bayes bounds.<\/jats:p>","DOI":"10.1561\/2200000100","type":"journal-article","created":{"date-parts":[[2024,1,22]],"date-time":"2024-01-22T08:17:48Z","timestamp":1705911468000},"page":"174-303","source":"Crossref","is-referenced-by-count":55,"title":["User-friendly Introduction to PAC-Bayes Bounds"],"prefix":"10.1108","volume":"17","author":[{"given":"Pierre","family":"Alquier","sequence":"first","affiliation":[{"name":"ESSEC Business School , Asia-Pacific Campus,","place":["Singapore"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"140","published-online":{"date-parts":[[2024,1,22]]},"reference":[{"key":"2026033012315573400_ref001","first-page":"6","volume-title":"Transductive and inductive adaptative inference for regression and density estimation","author":"Alquier","year":"2006"},{"issue":"4","key":"2026033012315573400_ref002","doi-asserted-by":"crossref","first-page":"279","DOI":"10.3103\/S1066530708040017","article-title":"PAC-Bayesian bounds for randomized empirical risk minimizers","volume":"17","author":"Alquier","year":"2008","journal-title":"Mathematical Methods of Statistics"},{"key":"2026033012315573400_ref003","first-page":"309","article-title":"Bayesian methods for low-rank matrix estimation: short survey and theoretical study","author":"Alquier","year":"2013"},{"issue":"1","key":"2026033012315573400_ref004","article-title":"Sparse single-index model","volume":"14","author":"Alquier","year":"2013","journal-title":"Journal of Machine Learning Research"},{"issue":"5","key":"2026033012315573400_ref005","doi-asserted-by":"crossref","first-page":"887","DOI":"10.1007\/s10994-017-5690-0","article-title":"Simpler PAC-Bayesian bounds for hostile data","volume":"107","author":"Alquier","year":"2018","journal-title":"Machine Learning"},{"issue":"2013","key":"2026033012315573400_ref006","doi-asserted-by":"crossref","first-page":"65","DOI":"10.2478\/demo-2013-0004","article-title":"Prediction of time series by statistical learning: general losses and fast rates","volume":"1","author":"Alquier","year":"2013","journal-title":"Dependence Modeling"},{"key":"2026033012315573400_ref007","doi-asserted-by":"crossref","first-page":"127","DOI":"10.1214\/11-EJS601","article-title":"PAC-Bayesian bounds for sparse regression estimation with exponential weights","volume":"5","author":"Alquier","year":"2011","journal-title":"Electronic Journal of Statistics"},{"issue":"3","key":"2026033012315573400_ref008","doi-asserted-by":"crossref","first-page":"1475","DOI":"10.1214\/19-AOS1855","article-title":"Concentration of tempered posteriors and of their variational approximations","volume":"48","author":"Alquier","year":"2020","journal-title":"Annals of Statistics"},{"issue":"239","key":"2026033012315573400_ref009","first-page":"1","article-title":"On the properties of variational approximations of Gibbs posteriors","volume":"17","author":"Alquier","year":"2016","journal-title":"Journal of Machine Learning Research"},{"issue":"3","key":"2026033012315573400_ref010","doi-asserted-by":"crossref","first-page":"883","DOI":"10.3150\/11-BEJ359","article-title":"Model selection for weakly dependent time series forecasting","volume":"18","author":"Alquier","year":"2012","journal-title":"Bernoulli"},{"key":"2026033012315573400_ref011","volume-title":"Advances in Neural Information Processing Systems","author":"Ambroladze","year":"2006"},{"key":"2026033012315573400_ref012","volume-title":"Characterizing the deneralization error of Gibbs algorithm with symmetrized KL information","author":"Aminian","year":"2021"},{"key":"2026033012315573400_ref013","first-page":"205","article-title":"Meta-learning by adjusting priors based on extended PAC-Bayes theory","author":"Amit","year":"2018"},{"key":"2026033012315573400_ref014","first-page":"9841017","article-title":"Exponential Smoothing for Off-Policy Learning","author":"Aouali","year":"2023"},{"key":"2026033012315573400_ref015","volume-title":"New bounds for k-means and information k-means","author":"Appert","year":"2021"},{"key":"2026033012315573400_ref016","volume-title":"Advances in Neural Information Processing Systems","author":"Asadi","year":"2018"},{"key":"2026033012315573400_ref017","volume-title":"PAC-Bayesian statistical learning theory","author":"Audibert","year":"2004"},{"issue":"4","key":"2026033012315573400_ref018","doi-asserted-by":"crossref","first-page":"1591","DOI":"10.1214\/08-AOS623","article-title":"Fast learning rates in statistical inference through aggregation","volume":"37","author":"Audibert","year":"2009","journal-title":"The Annals of Statistics"},{"issue":"4","key":"2026033012315573400_ref019","article-title":"Combining PAC-Bayesian and generic chaining bounds","volume":"8","author":"Audibert","year":"2007","journal-title":"Journal of Machine Learning Research"},{"issue":"5","key":"2026033012315573400_ref020","doi-asserted-by":"crossref","first-page":"2766","DOI":"10.1214\/11-AOS918","article-title":"Robust linear least squares regression","volume":"39","author":"Audibert","year":"2011","journal-title":"The Annals of Statistics"},{"key":"2026033012315573400_ref021","volume-title":"On the robustness to misspecification of a-posteriors and their variational approximations","author":"Avena Medina","year":"2021"},{"key":"2026033012315573400_ref022","first-page":"81","article-title":"On Bayesian bounds","author":"Banerjee","year":"2006"},{"issue":"3","key":"2026033012315573400_ref023","doi-asserted-by":"crossref","first-page":"313","DOI":"10.3390\/e23030313","article-title":"PAC-Bayes bounds on variational tempered posteriors for Markov models","volume":"23","author":"Banerjee","year":"2021","journal-title":"Entropy"},{"key":"2026033012315573400_ref024","first-page":"676","article-title":"Information complexity and generalization bounds","author":"Banerjee","year":"2021"},{"key":"2026033012315573400_ref025","volume-title":"Bayesian inference in high-dimensional models","author":"Banerjee","year":"2021"},{"issue":"6","key":"2026033012315573400_ref026","doi-asserted-by":"crossref","first-page":"2743","DOI":"10.1109\/18.720554","article-title":"The minimum description length principle in coding and modeling","volume":"44","author":"Barron","year":"1998","journal-title":"IEEE Transactions on Information Theory"},{"issue":"473","key":"2026033012315573400_ref027","doi-asserted-by":"crossref","first-page":"138","DOI":"10.1198\/016214505000000907","article-title":"Convexity, classification, and risk bounds","volume":"101","author":"Bartlett","year":"2006","journal-title":"Journal of the American Statistical Association"},{"issue":"3","key":"2026033012315573400_ref028","doi-asserted-by":"crossref","first-page":"311","DOI":"10.1007\/s00440-005-0462-3","article-title":"Empirical minimization","volume":"135","author":"Bartlett","year":"2006","journal-title":"Probability theory and related fields"},{"key":"2026033012315573400_ref029","first-page":"101","article-title":"Prediction and limiting synthesis of recursively enumerable classes of functions","volume":"210","author":"Barzdin\u0161","year":"1974","journal-title":"Latvijas Valsts Univ. Zimatm. Raksti"},{"key":"2026033012315573400_ref030","first-page":"25","article-title":"Learners that use little information","author":"Bassily","year":"2018"},{"key":"2026033012315573400_ref031","first-page":"435","volume-title":"Artificial Intel ligence and Statistics","author":"B\u00e9gin","year":"2016"},{"issue":"1","key":"2026033012315573400_ref032","doi-asserted-by":"crossref","first-page":"39","DOI":"10.1214\/18-AOS1712","article-title":"Bayesian fractional posteriors","volume":"47","author":"Bhattacharya","year":"2019","journal-title":"The Annals of Statistics"},{"issue":"10","key":"2026033012315573400_ref033","doi-asserted-by":"crossref","DOI":"10.3390\/e23101280","article-title":"Differentiable PAC-Bayes objectives with partially aggregated neural networks","volume":"23","author":"Biggs","year":"2021","journal-title":"Entropy"},{"key":"2026033012315573400_ref034","first-page":"3709","article-title":"On margins and derandomisation in PAC-Bayes","author":"Biggs","year":"2022"},{"key":"2026033012315573400_ref035","volume-title":"Minimax rates for conditional density estimation via empirical entropy","author":"Bilodeau","year":"2021"},{"issue":"5","key":"2026033012315573400_ref036","doi-asserted-by":"crossref","first-page":"1103","DOI":"10.1111\/rssb.12158","article-title":"A general framework for updating belief distributions","volume":"78","author":"Bissiri","year":"2016","journal-title":"Journal of the Royal Statistical Society: Series B (Statistical Methodology)"},{"key":"2026033012315573400_ref037","first-page":"112","article-title":"Occam\u2019s hammer","author":"Blanchard","year":"2007"},{"issue":"518","key":"2026033012315573400_ref038","doi-asserted-by":"crossref","first-page":"859","DOI":"10.1080\/01621459.2017.1285773","article-title":"Variational inference: A review for statisticians","volume":"112","author":"Blei","year":"2017","journal-title":"Journal of the American statistical Association"},{"key":"2026033012315573400_ref039","doi-asserted-by":"crossref","DOI":"10.1093\/acprof:oso\/9780199535255.001.0001","volume-title":"Concentration inequalities","author":"Boucheron","year":"2013"},{"issue":"1","key":"2026033012315573400_ref040","doi-asserted-by":"crossref","first-page":"121","DOI":"10.1109\/JSAIT.2020.2991139","article-title":"Tightening mutual information-based bounds on generalization error","volume":"1","author":"Bu","year":"2020","journal-title":"IEEE Journal on Selected Areas in Information Theory"},{"issue":"1","key":"2026033012315573400_ref041","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1561\/2200000024","article-title":"Regret analysis of stochastic and nonstochastic multi-armed bandit problems","volume":"5","author":"Bubeck","year":"2012","journal-title":"Foundations and Trends\u00ae in Machine Learning"},{"issue":"4","key":"2026033012315573400_ref042","doi-asserted-by":"crossref","first-page":"1725","DOI":"10.1109\/TIT.2008.917657","article-title":"Sequential procedures for aggregating arbitrary estimators of a conditional mean","volume":"54","author":"Bunea","year":"2008","journal-title":"IEEE Transactions on Information Theory"},{"key":"2026033012315573400_ref043","volume-title":"PAC-Bayesian supervised classification: The thermodynamics of statistical learning. Institute of Mathematical Statistics Lecture Notes - Monograph Series, 56","author":"Catoni","year":"2007"},{"key":"2026033012315573400_ref044","article-title":"A PAC-Bayesian approach to adaptive classification","volume-title":"Preprint LPMA 840","author":"Catoni","year":"2003"},{"key":"2026033012315573400_ref045","first-page":"1","volume-title":"Statistical learning theory and stochastic optimization. Saint-Flour Summer School on Probability Theory 2001 (Jean Picard ed.), Lecture Notes in Mathematics","author":"Catoni","year":"2004"},{"issue":"4","key":"2026033012315573400_ref046","first-page":"1148","article-title":"Challenging the empirical mean and empirical variance: a deviation study","volume":"48","author":"Catoni","year":"2012","journal-title":"Annales de l\u2019IHP Probabilit\u00e9s et statistiques"},{"key":"2026033012315573400_ref047","article-title":"Dimension free PAC-Bayesian bounds for the estimation of the mean of a random vector","author":"Catoni","year":"2017"},{"issue":"3","key":"2026033012315573400_ref048","doi-asserted-by":"crossref","first-page":"427","DOI":"10.1145\/258128.258179","article-title":"How to use expert advice","volume":"44","author":"Cesa-Bianchi","year":"1997","journal-title":"Journal of the ACM"},{"key":"2026033012315573400_ref049","doi-asserted-by":"crossref","DOI":"10.1017\/CBO9780511546921","volume-title":"Prediction, learning, and games","author":"Cesa-Bianchi","year":"2006"},{"key":"2026033012315573400_ref050","volume-title":"Interactions between compressed sensing random matrices and high dimensional geometry","author":"Chafa\u00ef","year":"2012"},{"key":"2026033012315573400_ref051","volume-title":"Learning with BOT-Bregman and Optimal Transport divergences","author":"Chee","year":"2021"},{"key":"2026033012315573400_ref052","first-page":"11","article-title":"Consistency of ELBO maximization for model selection","author":"Cherief-Abdellatif","year":"2019"},{"key":"2026033012315573400_ref053","first-page":"1831","article-title":"Convergence rates of variational inference in sparse deep learning","author":"Ch\u00e9rief-Abdellatif","year":"2020"},{"issue":"2","key":"2026033012315573400_ref054","doi-asserted-by":"crossref","first-page":"2995","DOI":"10.1214\/18-EJS1475","article-title":"Consistency of variational Bayes inference for estimation and model selection in mixtures","volume":"12","author":"Ch\u00e9rief-Abdellatif","year":"2018","journal-title":"Electronic Journal of Statistics"},{"key":"2026033012315573400_ref055","first-page":"662","article-title":"A generalization bound for online variational inference","author":"Ch\u00e9rief-Abdellatif","year":"2019"},{"key":"2026033012315573400_ref056","first-page":"3066","article-title":"On PAC-Bayesian reconstruction guarantees for VAEs","author":"Ch\u00e9rief-Abdellatif","year":"2022"},{"key":"2026033012315573400_ref057","first-page":"293","article-title":"On some recent advances on high dimensional Bayesian statistics","author":"Chopin","year":"2015"},{"key":"2026033012315573400_ref058","volume-title":"A unified recipe for deriving (time-uniform) PAC-Bayes bounds","author":"Chugg","year":"2023"},{"key":"2026033012315573400_ref059","first-page":"2311","article-title":"Conditionally Gaussian PAC-Bayes","author":"Clerico","year":"2022"},{"key":"2026033012315573400_ref060","volume-title":"A PAC- Bayes bound for deterministic classifiers","author":"Clerico","year":"2022"},{"key":"2026033012315573400_ref061","first-page":"4212","article-title":"Chained generalisation bounds","author":"Clerico","year":"2022"},{"key":"2026033012315573400_ref062","first-page":"447","article-title":"Wide stochastic networks: Gaussian limit and PAC-Bayesian training","author":"Clerico","year":"2023"},{"issue":"3","key":"2026033012315573400_ref063","doi-asserted-by":"crossref","first-page":"579","DOI":"10.1007\/s10994-017-5667-z","article-title":"1-bit matrix completion: PAC- Bayesian analysis of a variational approximation","volume":"107","author":"Cottet","year":"2018","journal-title":"Machine Learning"},{"issue":"1","key":"2026033012315573400_ref064","doi-asserted-by":"crossref","first-page":"302","DOI":"10.1214\/14-EJS886","article-title":"Aggregation of affine estimators","volume":"8","author":"Dai","year":"2014","journal-title":"Electronic Journal of Statistics"},{"issue":"3","key":"2026033012315573400_ref065","doi-asserted-by":"crossref","first-page":"18781905","DOI":"10.1214\/12-AOS1025","article-title":"Deviation optimal learning using greedy Q-aggregation","volume":"40","author":"Dai","year":"2012","journal-title":"The Annals of Statistics"},{"issue":"5","key":"2026033012315573400_ref066","doi-asserted-by":"crossref","first-page":"2452","DOI":"10.1214\/17-AOS1626","article-title":"On the exponentially weighted aggregate with the Laplace prior","volume":"46","author":"Dalalyan","year":"2018","journal-title":"Annals of Statistics"},{"issue":"4","key":"2026033012315573400_ref067","doi-asserted-by":"crossref","first-page":"2327","DOI":"10.1214\/12-AOS1038","article-title":"Sharp oracle inequalities for aggregation of affine estimators","volume":"40","author":"Dalalyan","year":"2012","journal-title":"The Annals of Statistics"},{"issue":"1-2","key":"2026033012315573400_ref068","doi-asserted-by":"crossref","first-page":"39","DOI":"10.1007\/s10994-008-5051-0","article-title":"Aggregation by exponential weighting, sharp PAC-Bayesian bounds and sparsity","volume":"72","author":"Dalalyan","year":"2008","journal-title":"Machine Learning"},{"issue":"5","key":"2026033012315573400_ref069","doi-asserted-by":"crossref","first-page":"1423","DOI":"10.1016\/j.jcss.2011.12.023","article-title":"Sparse regression learning by aggregation and Langevin Monte-Carlo","volume":"78","author":"Dalalyan","year":"2012","journal-title":"Journal of Computer and System Sciences"},{"issue":"2","key":"2026033012315573400_ref070","doi-asserted-by":"crossref","first-page":"1465","DOI":"10.1214\/19-AIHP1010","article-title":"Exponential weights in multivariate regression and a low-rankness favoring prior","volume":"56","author":"Dalalyan","year":"2020","journal-title":"Ann. Inst. H. Poincar\u00e9 Probab. Statist"},{"key":"2026033012315573400_ref071","doi-asserted-by":"crossref","first-page":"9","DOI":"10.1007\/978-0-387-69952-3_2","volume-title":"Weak dependence: With examples and applications","author":"Dedecker","year":"2007"},{"key":"2026033012315573400_ref072","doi-asserted-by":"crossref","DOI":"10.1007\/978-1-4612-0711-5","volume-title":"A probabilistic theory of pattern recognition","author":"Devroye","year":"1996"},{"key":"2026033012315573400_ref073","doi-asserted-by":"crossref","first-page":"389","DOI":"10.1002\/cpa.3160290405","article-title":"Asymptotic evaluation of certain Markov process expectations for large time. III","volume":"28","author":"Donsker","year":"1976","journal-title":"Communications on Pure and Applied Mathematics"},{"key":"2026033012315573400_ref074","first-page":"604","article-title":"On the role of data in PAC-Bayes bounds","author":"Dziugaite","year":"2021"},{"key":"2026033012315573400_ref075","article-title":"Computing nonvacuous generalization bounds for deep (stochastic) neural networks with many more parameters than training data","author":"Dziugaite","year":"2017"},{"key":"2026033012315573400_ref076","first-page":"8430","article-title":"Data-dependent PAC-Bayes priors via differential privacy","volume-title":"Advances in Neural Information Processing Systems","author":"Dziugaite","year":"2018"},{"key":"2026033012315573400_ref077","first-page":"604","article-title":"On the role of data in PAC-Bayes","author":"Dziugaite","year":"2021"},{"key":"2026033012315573400_ref078","first-page":"6626","volume-title":"PAC-Bayesian theory for stochastic LTI systems","author":"Eringis","year":"2021"},{"key":"2026033012315573400_ref079","unstructured":"Fleuret, F.\n           (2011). \u201cMachine learning, PAC-learning\u201d. Slides available on the author\u2019s website. URL: https:\/\/fleuret.org\/public\/EN_20110511-pac\/pac-fleuret-2011.pdf."},{"key":"2026033012315573400_ref080","first-page":"4093","volume-title":"Advances in Neural Information Processing Systems","author":"Foong","year":"2021"},{"key":"2026033012315573400_ref081","volume-title":"Sharpness- aware minimization for efficiently improving generalization","author":"Foret","year":"2020"},{"key":"2026033012315573400_ref082","volume-title":"Loss-based variational Bayes prediction","author":"Frazier","year":"2021"},{"key":"2026033012315573400_ref083","doi-asserted-by":"crossref","first-page":"538","DOI":"10.1214\/07-EJS077","article-title":"Optimal rates and adaptation in the single-index model using aggregation","volume":"1","author":"Ga\u00efffas","year":"2007","journal-title":"Electronic journal of statistics"},{"key":"2026033012315573400_ref084","volume-title":"On the difficulty of unbiased alpha divergence minimization","author":"Geffner","year":"2020"},{"key":"2026033012315573400_ref085","first-page":"1884","article-title":"PAC- Bayesian theory meets Bayesian inference","author":"Germain","year":"2016"},{"key":"2026033012315573400_ref086","first-page":"859","article-title":"A new PAC-Bayesian perspective on domain adaptation","author":"Germain","year":"2016"},{"issue":"26","key":"2026033012315573400_ref087","first-page":"787","article-title":"Risk bounds for the majority vote: from a PAC-Bayesian analysis to a learning algorithm","volume":"16","author":"Germain","year":"2015","journal-title":"Journal of Machine Learning Research"},{"key":"2026033012315573400_ref088","first-page":"353","article-title":"PAC- Bayesian learning of linear classifiers","author":"Germain","year":"2009"},{"key":"2026033012315573400_ref089","doi-asserted-by":"crossref","DOI":"10.1017\/9781139029834","volume-title":"Fundamentals of nonparametric Bayesian inference","author":"Ghosal","year":"2017"},{"key":"2026033012315573400_ref090","doi-asserted-by":"crossref","DOI":"10.1201\/b17895","volume-title":"Introduction to high-dimensional statistics","author":"Giraud","year":"2014"},{"key":"2026033012315573400_ref091","doi-asserted-by":"crossref","first-page":"3864","DOI":"10.3150\/17-BEJ981","article-title":"Robust dimension-free Gram operator estimates","volume":"24","author":"Giulini","year":"2018","journal-title":"Bernoulli"},{"key":"2026033012315573400_ref092","volume-title":"Bayes complexity of learners vs overfitting","author":"Gluch","year":"2023"},{"key":"2026033012315573400_ref093","first-page":"2217","article-title":"PAC-Bayes, MAC-Bayes and conditional mutual information: fast rate bounds that handle general VC classes","author":"Gr\u00fcnwald","year":"2021"},{"issue":"4","key":"2026033012315573400_ref094","doi-asserted-by":"crossref","first-page":"1069","DOI":"10.1214\/17-BA1085","article-title":"Inconsistency of Bayesian inference for misspecified linear models, and a proposal for repairing it","volume":"12","author":"Gr\u00fcnwald","year":"2017","journal-title":"Bayesian Analysis"},{"key":"2026033012315573400_ref095","doi-asserted-by":"crossref","DOI":"10.7551\/mitpress\/4643.001.0001","volume-title":"The minimum description length principle","author":"Gr\u00fcnwald","year":"2007"},{"issue":"56","key":"2026033012315573400_ref096","first-page":"1","article-title":"Fast rates for general unbounded loss functions: From ERM to Generalized Bayes","volume":"21","author":"Gr\u00fcnwald","year":"2020","journal-title":"Journal of Machine Learning Research"},{"key":"2026033012315573400_ref097","article-title":"A primer on PAC-Bayesian learning","author":"Guedj","year":"2019"},{"key":"2026033012315573400_ref098","doi-asserted-by":"crossref","first-page":"264","DOI":"10.1214\/13-EJS771","article-title":"PAC-Bayesian estimation and prediction in sparse additive models","volume":"7","author":"Guedj","year":"2013","journal-title":"Electronic Journal of Statistics"},{"key":"2026033012315573400_ref099","first-page":"25725","volume-title":"Advances in Neural Information Processing Systems","author":"Haddouche","year":"2022"},{"key":"2026033012315573400_ref100","article-title":"PAC-Bayes generalisation bounds for heavy-tailed losses through supermartingales","volume-title":"Transactions on Machine Learning Research","author":"Haddouche","year":"2023"},{"key":"2026033012315573400_ref101","volume-title":"Upper and lower bounds on the performance of kernel PCA","author":"Haddouche","year":"2020"},{"issue":"10","key":"2026033012315573400_ref102","doi-asserted-by":"crossref","DOI":"10.3390\/e23101330","article-title":"PAC-Bayes unleashed: generalisation bounds with unbounded losses","volume":"23","author":"Haddouche","year":"2021","journal-title":"Entropy"},{"key":"2026033012315573400_ref103","first-page":"26370","volume-title":"Advances in Neural Information Processing Systems","author":"Haghifam","year":"2021"},{"key":"2026033012315573400_ref104","first-page":"2487","article-title":"Understanding generalization via leave-one-out conditional mutual information","author":"Haghifam","year":"2022"},{"key":"2026033012315573400_ref105","first-page":"9925","volume-title":"Advances in Neural Information Processing Systems","author":"Haghifam","year":"2020"},{"key":"2026033012315573400_ref106","first-page":"663","article-title":"Limitations of information- theoretic generalization bounds for gradient descent methods in stochastic convex optimization","author":"Haghifam","year":"2023"},{"key":"2026033012315573400_ref107","first-page":"478","article-title":"Learning partially known stochastic dynamics with empirical PAC-Bayes","author":"Hau\u00dfmann","year":"2021"},{"issue":"3","key":"2026033012315573400_ref108","doi-asserted-by":"crossref","first-page":"824","DOI":"10.1109\/JSAIT.2020.3040992","article-title":"Generalization bounds via information density and conditional information density","volume":"1","author":"Hellstr\u00f6m","year":"2020","journal-title":"IEEE Journal on Selected Areas in Information Theory"},{"issue":"12","key":"2026033012315573400_ref109","doi-asserted-by":"crossref","first-page":"3140","DOI":"10.1109\/TIT.2002.805090","article-title":"A PAC-Bayesian margin bound for linear classifiers","volume":"48","author":"Herbrich","year":"2002","journal-title":"IEEE Transactions on Information Theory"},{"key":"2026033012315573400_ref110","first-page":"148","article-title":"A PAC-Bayes bound for tailored density estimation","author":"Higgs","year":"2010"},{"key":"2026033012315573400_ref111","first-page":"5","article-title":"Keeping the neural networks simple by minimizing the description length of the weights","author":"Hinton","year":"1993"},{"key":"2026033012315573400_ref112","first-page":"2067","article-title":"The many faces of exponential weights in online learning","author":"Hoeven","year":"2018"},{"key":"2026033012315573400_ref113","first-page":"2715","article-title":"PAC-Bayes under potentially heavy tails","volume":"32","author":"Holland","year":"2019","journal-title":"Advances in Neural Information Processing Systems"},{"key":"2026033012315573400_ref114","first-page":"384","article-title":"Tight bounds for the expected risk of linear classifiers and PAC-Bayes finite-sample guarantees","author":"Honorio","year":"2014"},{"key":"2026033012315573400_ref115","volume-title":"Practical bounds on the error of Bayesian posterior approximations: A nonasymptotic approach","author":"Huggins","year":"2018"},{"issue":"156","key":"2026033012315573400_ref116","first-page":"1","article-title":"Asymptotic consistency of a-R\u00e9nyi-approximate posteriors","volume":"21","author":"Jaiswal","year":"2020","journal-title":"Journal of Machine Learning Research"},{"key":"2026033012315573400_ref117","first-page":"2240","article-title":"Tighter PAC-Bayes Bounds Through Coin-Betting","author":"Jang","year":"2023"},{"key":"2026033012315573400_ref118","first-page":"2207","article-title":"Gibbs posterior for variable selection in high-dimensional classification and data mining","volume-title":"The Annals of Statistics","author":"Jiang","year":"2008"},{"key":"2026033012315573400_ref119","volume-title":"Weight expansion: a new perspective on dropout and generalization","author":"Jin","year":"2022"},{"issue":"1","key":"2026033012315573400_ref120","doi-asserted-by":"crossref","first-page":"474","DOI":"10.1109\/TIT.2021.3119605","article-title":"Transfer meta-learning: Information-theoretic bounds and information meta-risk minimization","volume":"68","author":"Jose","year":"2021","journal-title":"IEEE Transactions on Information Theory"},{"key":"2026033012315573400_ref121","first-page":"681","article-title":"Functional aggregation for nonparametric regression","volume-title":"Annals of Statistics","author":"Juditsky","year":"2000"},{"issue":"5","key":"2026033012315573400_ref122","doi-asserted-by":"crossref","first-page":"2183","DOI":"10.1214\/07-AOS546","article-title":"Learning by mirror averaging","volume":"36","author":"Juditsky","year":"2008","journal-title":"The Annals of Statistics"},{"key":"2026033012315573400_ref123","volume-title":"Advances in Neural Information Processing Systems","author":"Kakade","year":"2008"},{"key":"2026033012315573400_ref124","volume-title":"The Bayesian Learning Rule","author":"Khan","year":"2021"},{"key":"2026033012315573400_ref125","first-page":"153","article-title":"Averaging expert predictions","author":"Kivinen","year":"1999"},{"issue":"132","key":"2026033012315573400_ref126","first-page":"1","article-title":"An optimizationcentric view on Bayes\u2019 rule: Reviewing and generalizing variational inference","volume":"23","author":"Knoblauch","year":"2022","journal-title":"Journal of Machine Learning Research"},{"key":"2026033012315573400_ref127","volume-title":"Information theory and statistics","author":"Kullback","year":"1959"},{"key":"2026033012315573400_ref128","volume-title":"Advances in Neural Information Processing Systems","author":"Lacasse","year":"2006"},{"key":"2026033012315573400_ref129","volume-title":"PAC-Bayesian generalization bounds for multiLayer perceptrons","author":"Lan","year":"2020"},{"issue":"2","key":"2026033012315573400_ref130","doi-asserted-by":"crossref","first-page":"165","DOI":"10.1023\/A:1022806918936","article-title":"Microchoice bounds and selfbounding learning algorithms","volume":"51","author":"Langford","year":"2003","journal-title":"Machine Learning"},{"key":"2026033012315573400_ref131","first-page":"809","article-title":"(Not) bounding the true error","volume":"2","author":"Langford","year":"2002","journal-title":"Advances in Neural Information Processing Systems"},{"key":"2026033012315573400_ref132","volume-title":"Bounds for averaging classifiers","author":"Langford","year":"2001"},{"key":"2026033012315573400_ref133","first-page":"439","article-title":"PAC-Bayes & margins","author":"Langford","year":"2002"},{"key":"2026033012315573400_ref134","first-page":"5","article-title":"From PAC-Bayes bounds to quadratic programs for majority votes","author":"Laviolette","year":"2011"},{"key":"2026033012315573400_ref135","volume-title":"Aggregation procedures: optimality and fast rates","author":"Lecu\u00e9","year":"2007"},{"issue":"2","key":"2026033012315573400_ref136","doi-asserted-by":"crossref","first-page":"646","DOI":"10.3150\/11-BEJ408","article-title":"On the optimality of the aggregate with exponential weights for low temperatures","volume":"19","author":"Lecu\u00e9","year":"2013","journal-title":"Bernoulli"},{"issue":"4","key":"2026033012315573400_ref137","doi-asserted-by":"crossref","first-page":"682","DOI":"10.1137\/1136085","article-title":"Asymptotically minimax adaptive estimation I: upper bounds","volume":"36","author":"Lepski","year":"1992","journal-title":"Theory of Probability and its Applications"},{"key":"2026033012315573400_ref138","first-page":"6872","article-title":"Dichotomize and generalize: PAC-Bayesian binary activated deep neural networks","volume-title":"Advances in Neural Information Processing Systems","author":"Letarte","year":"2019"},{"issue":"8","key":"2026033012315573400_ref139","doi-asserted-by":"crossref","first-page":"3396","DOI":"10.1109\/TIT.2006.878172","article-title":"Information theory and mixing least-squares regressions","volume":"52","author":"Leung","year":"2006","journal-title":"IEEE Trans. Inform. Theory"},{"key":"2026033012315573400_ref140","first-page":"119","article-title":"Distributiondependent PAC-Bayes Priors","author":"Lever","year":"2010"},{"key":"2026033012315573400_ref141","doi-asserted-by":"crossref","first-page":"4","DOI":"10.1016\/j.tcs.2012.10.013","article-title":"Tighter PAC- Bayes bounds through distribution-dependent priors","volume":"473","author":"Lever","year":"2013","journal-title":"Theoretical Computer Science"},{"key":"2026033012315573400_ref142","first-page":"256","article-title":"The weighted majority algorithm","author":"Littlestone","year":"1989"},{"key":"2026033012315573400_ref143","volume-title":"PAC-Bayes bounds for meta-learning with data-dependent prior","author":"Liu","year":"2021"},{"key":"2026033012315573400_ref144","doi-asserted-by":"crossref","first-page":"391","DOI":"10.1016\/j.neucom.2021.09.018","article-title":"Statistical generalization performance guarantee for meta-learning with data dependent prior","volume":"465","author":"Liu","year":"2021","journal-title":"Neurocomputing"},{"key":"2026033012315573400_ref145","first-page":"20543","volume-title":"Advances in Neural Information Processing Systems","author":"Livni","year":"2020"},{"key":"2026033012315573400_ref146","first-page":"2931","article-title":"A PAC-Bayesian analysis of randomized learning with application to stochastic gradient descent","volume-title":"Advances in Neural Information Processing Systems","author":"London","year":"2017"},{"key":"2026033012315573400_ref147","first-page":"4125","article-title":"Bayesian counterfactual risk minimization","author":"London","year":"2019"},{"issue":"8-9","key":"2026033012315573400_ref148","doi-asserted-by":"crossref","first-page":"1503","DOI":"10.1007\/s10994-019-05803-4","article-title":"On PAC-Bayesian bounds for random forests","volume":"108","author":"Lorenzen","year":"2019","journal-title":"Machine Learning"},{"key":"2026033012315573400_ref149","volume-title":"Online-to-PAC conversions: Generalization bounds via regret analysis","author":"Lugosi","year":"2021"},{"key":"2026033012315573400_ref150","first-page":"3524","article-title":"Generalization bounds via convex analysis","author":"Lugosi","year":"2022"},{"key":"2026033012315573400_ref151","doi-asserted-by":"crossref","first-page":"209","DOI":"10.1016\/j.jmva.2018.12.004","article-title":"PAC-Bayesian risk bounds for group-analysis sparse regression by exponential weighting","volume":"171","author":"Luu","year":"2019","journal-title":"Journal of Multivariate Analysis"},{"key":"2026033012315573400_ref152","volume-title":"PAC-Bayesian estimation of low-rank matrices","author":"Mai","year":"2017"},{"issue":"2","key":"2026033012315573400_ref153","doi-asserted-by":"crossref","first-page":"333","DOI":"10.3390\/e25020333","article-title":"From bilinear regression to inductive matrix completion: a quasi-Bayesian analysis","volume":"25","author":"Mai","year":"2023","journal-title":"Entropy"},{"key":"2026033012315573400_ref154","first-page":"1","article-title":"Simulation comparisons between Bayesian and de-biased estimators in low-rank matrix completion","volume-title":"METRON","author":"Mai","year":"2023"},{"issue":"1","key":"2026033012315573400_ref155","doi-asserted-by":"crossref","first-page":"823","DOI":"10.1214\/15-EJS1020","article-title":"A Bayesian approach for noisy matrix completion: Optimal rate under general sampling distribution","volume":"9","author":"Mai","year":"2015","journal-title":"Electronic Journal of Statistics"},{"key":"2026033012315573400_ref156","doi-asserted-by":"crossref","first-page":"62","DOI":"10.1016\/j.jspi.2016.11.003","article-title":"Pseudo-Bayesian quantum tomography with rank-adaptation","volume":"184","author":"Mai","year":"2017","journal-title":"Journal of Statistical Planning and Inference"},{"issue":"6","key":"2026033012315573400_ref157","doi-asserted-by":"crossref","first-page":"1808","DOI":"10.1214\/aos\/1017939240","article-title":"Smooth discrimination analysis","volume":"27","author":"Mammen","year":"1999","journal-title":"The Annals of Statistics"},{"key":"2026033012315573400_ref158","first-page":"5263","volume-title":"Advances in Neural Information Processing Systems","author":"Masegosa","year":"2020"},{"key":"2026033012315573400_ref159","first-page":"5479","volume-title":"Advances in Neural Information Processing Systems","author":"Masegosa","year":"2020"},{"key":"2026033012315573400_ref160","volume-title":"A note on the PAC Bayesian theorem","author":"Maurer","year":"2004"},{"key":"2026033012315573400_ref161","volume-title":"PAC-Bayesian generalization bounds for adversarial generative models","author":"Mbacke","year":"2023"},{"key":"2026033012315573400_ref162","first-page":"230","article-title":"Some PAC-Bayesian theorems","author":"McAllester","year":"1998"},{"key":"2026033012315573400_ref163","first-page":"164","article-title":"PAC-Bayesian model averaging","author":"McAllester","year":"1999"},{"issue":"1","key":"2026033012315573400_ref164","doi-asserted-by":"crossref","first-page":"5","DOI":"10.1023\/A:1021840411064","article-title":"PAC-Bayesian stochastic model selection","volume":"51","author":"McAllester","year":"2003","journal-title":"Machine Learning"},{"key":"2026033012315573400_ref165","volume-title":"A PAC-Bayesian tutorial with a dropout bound","author":"McAllester","year":"2013"},{"key":"2026033012315573400_ref166","doi-asserted-by":"crossref","first-page":"195","DOI":"10.1007\/978-3-662-12788-9_6","volume-title":"Probabilistic methods for algorithmic discrete mathematics","author":"McDiarmid","year":"1998"},{"key":"2026033012315573400_ref167","first-page":"839","article-title":"Generalization error bounds for Bayesian mixture algorithms","volume":"4","author":"Meir","year":"2003","journal-title":"Journal of Machine Learning Research"},{"issue":"10","key":"2026033012315573400_ref168","doi-asserted-by":"crossref","DOI":"10.3390\/e23101257","article-title":"Meta-strategy for learning tuning parameters with guarantees","volume":"23","author":"Meunier","year":"2021","journal-title":"Entropy"},{"key":"2026033012315573400_ref169","volume-title":"Advances in Neural Information Processing Systems","author":"Mhammedi","year":"2019"},{"key":"2026033012315573400_ref170","volume-title":"Local Risk Bounds for Statistical Aggregation","author":"Mourtada","year":"2023"},{"key":"2026033012315573400_ref171","first-page":"1547","article-title":"A direct sum result for the information complexity of learning","author":"Nachum","year":"2018"},{"key":"2026033012315573400_ref172","volume-title":"Dimension-free bounds for sum of dependent matrices and operators with heavytailed distribution","author":"Nakakita","year":"2022"},{"key":"2026033012315573400_ref173","first-page":"11015","article-title":"Information-theoretic generalization bounds for SGLD via data-dependent estimates","volume":"32","author":"Negrea","year":"2019","journal-title":"Advances in Neural Information Processing Systems"},{"key":"2026033012315573400_ref174","first-page":"85","article-title":"Topics in non-parametric statistics","volume":"28","author":"Nemirovski","year":"2000","journal-title":"Ecole d\u2019Et\u00e9 de Probabilit\u00e9s de Saint-Flour"},{"key":"2026033012315573400_ref175","first-page":"3526","article-title":"Information-theoretic generalization bounds for stochastic gradient descent","author":"Neu","year":"2021"},{"key":"2026033012315573400_ref176","article-title":"A PAC-Bayesian approach to spectrally-normalized margin bounds for neural networks","author":"Neyshabur","year":"2017"},{"key":"2026033012315573400_ref177","first-page":"21","article-title":"PAC-Bayesian contrastive unsupervised representation learning","author":"Nozawa","year":"2020"},{"key":"2026033012315573400_ref178","volume-title":"PAC-Bayes analysis of sentence representation","author":"Nozawa","year":"2019"},{"key":"2026033012315573400_ref179","volume-title":"Adaptive variational Bayes: Optimality, computation and applications","author":"Ohn","year":"2021"},{"key":"2026033012315573400_ref180","first-page":"1711","article-title":"Novel change of measure inequalities with applications to PAC-Bayesian bounds and Monte Carlo estimation","author":"Ohnishi","year":"2021"},{"key":"2026033012315573400_ref181","doi-asserted-by":"crossref","first-page":"231","DOI":"10.1016\/j.neucom.2019.12.137","article-title":"Randomized learning and generalization of fair and private classifiers: From PAC-Bayes to stability and differential privacy","volume":"416","author":"Oneto","year":"2020","journal-title":"Neurocomputing"},{"key":"2026033012315573400_ref182","volume-title":"A modern introduction to online learning","author":"Orabona","year":"2019"},{"issue":"1","key":"2026033012315573400_ref183","first-page":"3507","article-title":"PAC-Bayes bounds with data dependent priors","volume":"13","author":"Parrado-Hern\u00e2ndez","year":"2012","journal-title":"The Journal of Machine Learning Research"},{"key":"2026033012315573400_ref184","first-page":"991","article-title":"A PAC-Bayesian bound for lifelong learning","author":"Pentina","year":"2014"},{"issue":"1","key":"2026033012315573400_ref185","first-page":"10326","article-title":"Tighter risk certificates for neural networks","volume":"22","author":"P\u00e9rez-Ortiz","year":"2021","journal-title":"The Journal of Machine Learning Research"},{"key":"2026033012315573400_ref186","first-page":"7739","article-title":"Dissecting non-vacuous generalization bounds based on the mean-field approximation","author":"Pitas","year":"2020"},{"issue":"11","key":"2026033012315573400_ref187","doi-asserted-by":"crossref","DOI":"10.3390\/e22111263","article-title":"Dynamics of coordinate ascent variational inference: A case study in 2D Ising models","volume":"22","author":"Plummer","year":"2020","journal-title":"Entropy"},{"key":"2026033012315573400_ref188","first-page":"26","article-title":"Information-theoretic analysis of stability and bias of learning algorithms","author":"Raginsky","year":"2016"},{"key":"2026033012315573400_ref189","first-page":"1927","article-title":"Chromatic PAC-Bayes bounds for non-i.i.d. data: Applications to ranking and stationary \u03b2-mixing processes","volume":"11","author":"Ralaivola","year":"2010","journal-title":"Journal of Machine Learning Research"},{"key":"2026033012315573400_ref190","volume-title":"A general framework for PAC-Bayes bounds for meta-learning","author":"Rezazadeh","year":"2022"},{"key":"2026033012315573400_ref191","first-page":"658","article-title":"PAC-Bayesian AUC classification and scoring","volume":"1","author":"Ridgway","year":"2014","journal-title":"Advances in Neural Information Processing Systems"},{"issue":"10","key":"2026033012315573400_ref192","first-page":"905","article-title":"In\u00e9galit\u00e9s de Hoeffding pour les fonctions lipschitziennes de suites d\u00e9pendantes","volume":"330","author":"Rio","year":"2000","journal-title":"Comptes Rendus de l\u2019Acad\u00e9mie des Sciences- Series I-Mathematics"},{"key":"2026033012315573400_ref193","volume-title":"Bayes meets Bernstein at the meta level: an analysis of fast rates in meta-learning with PAC-Bayes","author":"Riou","year":"2023"},{"issue":"5","key":"2026033012315573400_ref194","doi-asserted-by":"crossref","first-page":"465","DOI":"10.1016\/0005-1098(78)90005-5","article-title":"Modeling by shortest data description","volume":"14","author":"Rissanen","year":"1978","journal-title":"Automat- ica"},{"key":"2026033012315573400_ref195","article-title":"PAC-Bayes analysis beyond the usual bounds","volume-title":"Advances in Neural Information Processing Systems","author":"Rivasplata","year":"2020"},{"key":"2026033012315573400_ref196","volume-title":"PAC-Bayes with backprop","author":"Rivasplata","year":"2019"},{"key":"2026033012315573400_ref197","volume-title":"More PAC-Bayes bounds: From bounded losses, to losses with general tail behaviors, to anytime-validity","author":"Rodrigues-Galvez","year":"2023"},{"key":"2026033012315573400_ref198","first-page":"19109","volume-title":"Advances in Neural Information Processing Systems","author":"Rodriguez-G\u00e0lvez","year":"2021"},{"key":"2026033012315573400_ref199","first-page":"9116","article-title":"PACOH: Bayes-optimal meta-learning with PAC-guarantees","author":"Rothfuss","year":"2021"},{"key":"2026033012315573400_ref200","doi-asserted-by":"crossref","first-page":"211","DOI":"10.1146\/annurev-statistics-041715-033523","article-title":"On the frequentist properties of Bayesian nonparametric methods","volume":"3","author":"Rousseau","year":"2016","journal-title":"Annual Review of Statistics and Its Application"},{"issue":"1","key":"2026033012315573400_ref201","doi-asserted-by":"crossref","first-page":"302","DOI":"10.1109\/TIT.2019.2945779","article-title":"How much does your data exploration overfit? Controlling bias via information usage","volume":"66","author":"Russo","year":"2019","journal-title":"IEEE Transactions on Information Theory"},{"key":"2026033012315573400_ref202","first-page":"29777","article-title":"PAC-Bayesian Offline Contextual Bandits With Guarantees","author":"Sakhi","year":"2023"},{"issue":"1","key":"2026033012315573400_ref203","doi-asserted-by":"crossref","first-page":"416","DOI":"10.1214\/aop\/1019160125","article-title":"Concentration of measure inequalities for Markov chains and \u03a6-mixing processes","volume":"28","author":"Samson","year":"2000","journal-title":"The Annals of Probability"},{"key":"2026033012315573400_ref204","first-page":"233","article-title":"PAC-Bayesian generalisation error bounds for Gaussian process classification","volume":"3","author":"Seeger","year":"2002","journal-title":"Journal of machine learning research"},{"key":"2026033012315573400_ref205","volume-title":"Tech. rep","author":"Seeger","year":"2003"},{"key":"2026033012315573400_ref206","first-page":"1683","article-title":"PAC-Bayesian analysis of contextual bandits","volume-title":"Advances in Neural Information Processing Systems","author":"Seldin","year":"2011"},{"key":"2026033012315573400_ref207","first-page":"98","article-title":"PAC-Bayes-Bernstein inequality for martingales and its application to multiarmed bandits","author":"Seldin","year":"2012"},{"issue":"12","key":"2026033012315573400_ref208","doi-asserted-by":"crossref","first-page":"7086","DOI":"10.1109\/TIT.2012.2211334","article-title":"PAC-Bayesian inequalities for martingales","volume":"58","author":"Seldin","year":"2012","journal-title":"IEEE Transactions on Information Theory"},{"issue":"12","key":"2026033012315573400_ref209","article-title":"PAC-Bayesian analysis of coclustering and beyond","volume":"11","author":"Seldin","year":"2010","journal-title":"Journal of Machine Learning Research"},{"issue":"2","key":"2026033012315573400_ref210","doi-asserted-by":"crossref","first-page":"107","DOI":"10.1561\/2200000018","article-title":"Online learning and online convex optimization","volume":"4","author":"Shalev-Shwartz","year":"2011","journal-title":"Foundations and Trends\u00ae in Machine Learning"},{"key":"2026033012315573400_ref211","first-page":"2","article-title":"A PAC analysis of a Bayes estimator","author":"Shawe-Taylor","year":"1997"},{"key":"2026033012315573400_ref212","first-page":"5151","article-title":"Excess risk bounds for the Bayes risk using variational inference in latent Gaussian models","volume-title":"Advances in Neural Information Processing Systems","author":"Sheth","year":"2017"},{"key":"2026033012315573400_ref213","volume-title":"PAC-Bayes training for neural networks: sparsity and uncertainty quantification","author":"Steffen","year":"2022"},{"key":"2026033012315573400_ref214","first-page":"3437","article-title":"Reasoning about generalization via conditional mutual information","author":"Steinke","year":"2020"},{"key":"2026033012315573400_ref215","first-page":"8145","article-title":"PAC-Bayesian learning of optimization algorithms","author":"Sucker","year":"2023"},{"key":"2026033012315573400_ref216","first-page":"8","article-title":"PAC-Bayesian bound for Gaussian process regression and multiple kernel additive model","author":"Suzuki","year":"2012"},{"key":"2026033012315573400_ref217","first-page":"1273","article-title":"Convergence rate of Bayesian tensor estimator and its minimax optimality","author":"Suzuki","year":"2015"},{"key":"2026033012315573400_ref218","first-page":"19224","volume-title":"Advances in Neural Information Processing Systems","author":"Suzuki","year":"2020"},{"issue":"2","key":"2026033012315573400_ref219","doi-asserted-by":"crossref","first-page":"479","DOI":"10.1093\/biomet\/asy054","article-title":"Calibrating general posterior credible regions","volume":"106","author":"Syring","year":"2019","journal-title":"Biometrika"},{"issue":"2","key":"2026033012315573400_ref220","doi-asserted-by":"crossref","first-page":"1080","DOI":"10.3150\/22-BEJ1491","article-title":"Gibbs posterior concentration rates under sub-exponential type losses","volume":"29","author":"Syring","year":"2023","journal-title":"Bernoulli"},{"key":"2026033012315573400_ref221","volume-title":"PAC-Bayesian soft actor-critic learning","author":"Tasdighi","year":"2023"},{"key":"2026033012315573400_ref222","first-page":"466","article-title":"A strongly quasiconvex PAC-Bayesian bound","author":"Thiemann","year":"2017"},{"key":"2026033012315573400_ref223","first-page":"1","article-title":"PAC-Bayes-empirical-Bernstein inequality","volume-title":"Advances in Neural Information Processing Systems 26 (NIPS 2013)","author":"Tolstikhin","year":"2013"},{"key":"2026033012315573400_ref224","first-page":"9636","article-title":"Normalized flat minima: Exploring scale invariant definition of flat minima for neural networks using PAC-Bayesian analysis","author":"Tsuzuku","year":"2020"},{"key":"2026033012315573400_ref225","first-page":"303","article-title":"Optimal rates of aggregation","author":"Tsybakov","year":"2003"},{"issue":"11","key":"2026033012315573400_ref226","doi-asserted-by":"crossref","first-page":"1134","DOI":"10.1145\/1968.1972","article-title":"A theory of the learnable","volume":"27","author":"Valiant","year":"1984","journal-title":"Communications of the ACM"},{"key":"2026033012315573400_ref227","volume-title":"PAC-Bayes mini-tutorial: a continuous union bound","author":"van Erven","year":"2014"},{"key":"2026033012315573400_ref228","volume-title":"Statistical learning theory","author":"Vapnik","year":"1998"},{"issue":"4","key":"2026033012315573400_ref229","first-page":"781","article-title":"The uniform convergence of frequencies of the appearance of events to their probabilities","volume":"181","author":"Vapnik","year":"1968","journal-title":"Doklady Akademii Nauk"},{"key":"2026033012315573400_ref230","article-title":"Interpreting neural networks as majority votes through the PAC-Bayesian theory","author":"Viallard","year":"2019"},{"key":"2026033012315573400_ref231","doi-asserted-by":"crossref","DOI":"10.1016\/B978-1-55860-146-8.50032-1","article-title":"Aggregating strategies","author":"Vovk","year":"1990"},{"key":"2026033012315573400_ref232","doi-asserted-by":"crossref","DOI":"10.1017\/9781108627771","volume-title":"High-dimensional statistics: A non-asymptotic viewpoint","author":"Wainwright","year":"2019"},{"key":"2026033012315573400_ref233","first-page":"6648","article-title":"On the generalization gap inreparameterizable reinforcement learning","author":"Wang","year":"2019"},{"key":"2026033012315573400_ref234","volume-title":"PAC-Bayes information bottleneck","author":"Wang","year":"2021"},{"key":"2026033012315573400_ref235","first-page":"489","article-title":"Deviation inequalities for sums of weakly dependent time series","volume-title":"Electronic Communications in Probability","author":"Wintenberger","year":"2010"},{"key":"2026033012315573400_ref236","first-page":"12625","volume-title":"Advances in Neural Information Processing Systems","author":"Wu","year":"2021"},{"key":"2026033012315573400_ref237","first-page":"11369","volume-title":"Advances in Neural Information Processing Systems","author":"Wu","year":"2022"},{"key":"2026033012315573400_ref238","first-page":"2521","article-title":"Information-theoretic analysis of generalization capability of learning algorithms","author":"Xu","year":"2017"},{"key":"2026033012315573400_ref239","first-page":"10803","article-title":"Fast-rate PAC-Bayes generalization bounds via shifted Rademacher processes","volume":"32","author":"Yang","year":"2019","journal-title":"Advances in Neural Information Processing Systems"},{"issue":"454","key":"2026033012315573400_ref240","doi-asserted-by":"crossref","first-page":"574","DOI":"10.1198\/016214501753168262","article-title":"Adaptive regression by mixing","volume":"96","author":"Yang","year":"2001","journal-title":"Journal of the American Statistical Association"},{"issue":"1","key":"2026033012315573400_ref241","first-page":"25","article-title":"Aggregating regression procedures to improve performance","volume":"10","author":"Yang","year":"2004","journal-title":"Bernoul li"},{"issue":"2","key":"2026033012315573400_ref242","doi-asserted-by":"crossref","first-page":"886","DOI":"10.1214\/19-AOS1827","article-title":"a -variational inference with statistical guarantees","volume":"48","author":"Yang","year":"2020","journal-title":"Annals of Statistics"},{"issue":"4","key":"2026033012315573400_ref243","doi-asserted-by":"crossref","first-page":"2180","DOI":"10.1214\/19-AOS1883","article-title":"Convergence rates of variational posterior distributions","volume":"48","author":"Zhang","year":"2020","journal-title":"Annals of Statistics"},{"issue":"4","key":"2026033012315573400_ref244","doi-asserted-by":"crossref","first-page":"1307","DOI":"10.1109\/TIT.2005.864439","article-title":"Information-theoretic upper and lower bounds for statistical estimation","volume":"52","author":"Zhang","year":"2006","journal-title":"IEEE Transactions on Information Theory"},{"key":"2026033012315573400_ref245","volume-title":"Auto-tune: PAC- Bayes optimization over prior and posterior for neural networks","author":"Zhang","year":"2023"},{"key":"2026033012315573400_ref246","volume-title":"Dimension-free bounds for sums of independent matrices and simple tensors via the variational principle","author":"Zhivotovskiy","year":"2021"},{"key":"2026033012315573400_ref247","volume-title":"Non-vacuous generalization bounds at the imagenet scale: a PAC- Bayesian compression approach","author":"Zhou","year":"2018"}],"container-title":["Foundations and Trends\u00ae in Machine Learning"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.emerald.com\/ftmal\/article-pdf\/17\/2\/174\/11156028\/2200000100en.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/www.emerald.com\/ftmal\/article-pdf\/17\/2\/174\/11156028\/2200000100en.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,29]],"date-time":"2026-04-29T18:10:52Z","timestamp":1777486252000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.emerald.com\/ftmal\/article\/17\/2\/174\/1332427\/User-friendly-Introduction-to-PAC-Bayes-Bounds"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,1,22]]},"references-count":247,"journal-issue":{"issue":"2","published-print":{"date-parts":[[2024,1,22]]}},"URL":"https:\/\/doi.org\/10.1561\/2200000100","relation":{},"ISSN":["1935-8237","1935-8245"],"issn-type":[{"value":"1935-8237","type":"print"},{"value":"1935-8245","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,1,22]]}}}