BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20260605T054449EDT-9713faEM8x@132.216.98.100
DTSTAMP:20260605T094449Z
DESCRIPTION:\n	\n		\n			\n				\n					\n						\n							\n								\n									\n										\n											\n												📍 LIEU / PLACE\n													Hybride - CRM\, Salle /
  Room 6214\, Pavillon André Aisenstadt\n													\n													✒️ TITRE / TITLE\n\n												Statistical 
 Frameworks for Trustworthy Machine Learning: Privacy\, Uncertainty\, and O
 nline Inference\n													\n													📄 RÉSUMÉ / ABSTRACT\n\n												Trustworthy machine learning req
 uires rigorous privacy protection and valid uncertainty quantification\, e
 specially in modern streaming settings. We present two complementary advan
 ces.\n													First\, we extend Gaussian Differential Privacy (GDP) to general Rie
 mannian manifolds. Leveraging the Bishop-Gromov comparison theorem\, we co
 nstruct a Riemannian Gaussian mechanism based on geodesic distance and cal
 ibrate the privacy parameter $\mu$ to achieve GDP under bounded Ricci curv
 ature. We provide practical calibration methods\, an efficient procedure o
 n one-dimensional manifolds and an MCMC-based algorithm on constant-curvat
 ure spaces. We demonstrate improved utility (e.g.\, on the sphere $S^d$) r
 elative to Riemannian Laplace mechanisms.\n													Second\, we develop online infe
 rence for smoothed quantile regression\, introducing an incremental updati
 ng estimator for low-dimensional models and an online debiased lasso for h
 igh-dimensional sparse settings. The procedures use only current data and 
 compact history summaries\, correct online approximation error\, and deliv
 er asymptotically valid confidence intervals and tests. Simulations and re
 al data analyses (e.g.\, bike-sharing demand and index-fund data) illustra
 te reliability and scalability.\n													Together\, these results provide privacy-
 preserving data access and statistically sound\, streaming-ready inference
 \, core ingredients of trustworthy machine learning.\n													\n													🍷 Une réception vi
 ns et fromages suivra / A wine and cheese reception will follow.\n											\n										\n									\n								\n
 							\n						\n					\n				\n			\n		\n	\n\n\n \n\n\n	\n		\n			\n				\n					\n						\n							\n								Lien ZOOM Link\n							\n						\n					\n				\n			\n		\n	\n\n
DTSTART:20251017T193000Z
DTEND:20251017T203000Z
SUMMARY:Linglong Kong (University of Alberta)
URL:https://www.mcgill.ca/mathstat/channels/event/linglong-kong-university-
 alberta-368399
END:VEVENT
END:VCALENDAR