BEGIN:VCALENDAR VERSION:2.0 PRODID:-//wp-events-plugin.com//7.4.0.1//EN TZID:Europe/Paris X-WR-TIMEZONE:Europe/Paris BEGIN:VEVENT UID:404@isdm.umontpellier.fr DTSTART;TZID=Europe/Paris:20241112T140000 DTEND;TZID=Europe/Paris:20241112T140000 DTSTAMP:20260828T115423Z URL:https://isdm.umontpellier.fr/events/model-free-conditional-conformal-d epth-measures-for-uncertainty-quantification-in-regression-models-in-separ able-hilbert-spaces-2/ SUMMARY:Model-Free Conditional Conformal Depth Measures for Uncertainty Qua ntification in Regression Models in Separable Hilbert Spaces DESCRIPTION:Online\n\nMachine Learning in Montpellier\, Theory &\; Pract ice\n\nDepth measures have gained popularity in the statistical literature for defining level sets in complex data structures such as multivariate d ata\, functional data\, and random graphs. Despite their versatility\, int egrating depth measures into regression modeling for establishing predicti on regions remains a largely underexplored area. To address this gap\, we propose novel model-free uncertainty quantification algorithms based on co nditional depth measures and the notion of conditional kernel mean embeddi ngs. The new algorithms can be used to define prediction and tolerance reg ions when the predictors and responses are defined in separable Hilbert sp aces. To enhance the practical utility of the algorithms\, we also introdu ce a conformal prediction version\, providing non-asymptotic guarantees fo r the derived prediction regions. Additionally\, we establish both conditi onal and unconditional consistency results and fast convergence rates in s ome special homoscedastic cases. We evaluate the model'\;s finite-samp le performance in extensive simulation studies involving different functio nal data objects and probability distributions. Finally\, we apply the app roach to a digital health application related to physical activity\, aimin g to offer personalized recommendations to the U.S. population based on in dividuals'\; characteristics.\n\nMachine Learning in Montpellier\, The ory &\; Practice ATTACH;FMTTYPE=image/jpeg:https://isdm.umontpellier.fr/wp-content/uploads/ 2026/06/ml-mtp-gC78d5.png CATEGORIES:ML MTP END:VEVENT BEGIN:VTIMEZONE TZID:Europe/Paris X-LIC-LOCATION:Europe/Paris BEGIN:STANDARD DTSTART:20241027T020000 TZOFFSETFROM:+0200 TZOFFSETTO:+0100 TZNAME:CET END:STANDARD END:VTIMEZONE END:VCALENDAR