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:359@isdm.umontpellier.fr
DTSTART;TZID=Europe/Paris:20250404T140000
DTEND;TZID=Europe/Paris:20250404T140000
DTSTAMP:20260828T094112Z
URL:https://isdm.umontpellier.fr/events/optimal-classification-under-perfo
 rmative-distribution-shift-3/
SUMMARY:Optimal Classification under Performative Distribution Shift
DESCRIPTION:Room 02.124\, Building 5\, St Priest campus\n\nMachine Learning
  in Montpellier\, Theory &amp\; Practice - Olivier Cappé ()\n\nPerformati
 ve learning addresses the increasingly pervasive situations in which algor
 ithmic decisions may induce changes in the data distribution as a conseque
 nce of their public deployment. We propose a novel view in which these per
 formative effects are modelled as push-forward measures. This general fram
 ework encompasses existing models and enables novel performative gradient 
 estimation methods\, leading to more efficient and scalable learning strat
 egies. For distribution shifts\, unlike previous models which require full
  specification of the data distribution\, we only assume knowledge of the 
 shift operator that represents the performative changes. Focusing on class
 ification with a linear-in-parameters performative effect\, we prove the c
 onvexity of the performative risk under a new set of assumptions. We also 
 establish a connection with adversarially robust classification by reformu
 lating the minimization of the performative risk as a min-max variational 
 problem.\n\nMachine Learning in Montpellier\, Theory &amp\; Practice
ATTACH;FMTTYPE=image/jpeg:https://isdm.umontpellier.fr/wp-content/uploads/
 2026/06/ml-mtp-gC78d5.png
CATEGORIES:ML MTP
LOCATION:Room 02.124 Building 5\, St Priest Campus\, Montpellier\, 34000\, 
 France
X-APPLE-STRUCTURED-LOCATION;VALUE=URI;X-ADDRESS=St Priest Campus\, Montpell
 ier\, 34000\, France;X-APPLE-RADIUS=100;X-TITLE=Room 02.124 Building 5:geo
 :0,0
END:VEVENT
BEGIN:VTIMEZONE
TZID:Europe/Paris
X-LIC-LOCATION:Europe/Paris
BEGIN:DAYLIGHT
DTSTART:20250330T030000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
END:DAYLIGHT
END:VTIMEZONE
END:VCALENDAR