BEGIN:VCALENDAR VERSION:2.0 PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4// BEGIN:VEVENT UID:20250921T175029EDT-9396mv88IV@132.216.98.100 DTSTAMP:20250921T215029Z DESCRIPTION:Title: Generalization bounds via regret analysis”\n\nAbstract: Understanding the generalization ability of learning algorithms has been a key driving force behind statistical learning theory.\n\nIn this talk\, w e present a novel framework for deriving bounds on the generalization erro r of statistical learning algorithms from the perspective of online learni ng. Specifically\, we construct an online learning game called the “genera lization game”\, where an online learner competes with a fixed statistical learning algorithm in predicting the sequence of generalization gaps on a training set of i.i.d. data points. We establish a connection between the online and statistical learning setting by showing that the existence of an online learning algorithm with bounded regret in this game implies a bo und on the generalization error of the statistical learning algorithm. Thi s technique allows us to recover several standard generalization bounds\, including a range of PAC-Bayesian and information-theoretic guarantees.\n DTSTART:20250922T190000Z DTEND:20250922T190000Z LOCATION:Room 1104\, Burnside Hall\, CA\, QC\, Montreal\, H3A 0B9\, 805 rue Sherbrooke Ouest SUMMARY:Gabor Lugosi (Universitat Pompeu Fabra) URL:/sustainability/channels/event/gabor-lugosi-univer sitat-pompeu-fabra-367890 END:VEVENT END:VCALENDAR