BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20260415T071951EDT-4360IV1wsx@132.216.98.100
DTSTAMP:20260415T111951Z
DESCRIPTION:A variational perspective for accelerated methods in optimizati
 on\n	Abstract:\n	Accelerated gradient methods play a central role in optimiz
 ation\, achieving optimal rates in many settings. While many generalizatio
 ns and extensions of Nesterov's original acceleration method have been pro
 posed\, it is not yet clear what is the natural scope of the acceleration 
 concept. In this work\, we study accelerated methods from a continuous-tim
 e perspective. We show that there is a Lagrangian functional that we call 
 the “Bregman Lagrangian” which generates a large class of accelerated meth
 ods in continuous time\, including (but not limited to) accelerated gradie
 nt descent\, its non-Euclidean extension\, and accelerated higher-order gr
 adient methods. We show that the continuous-time limit of all of these met
 hods correspond to traveling the same curve in spacetime at different spee
 ds. From this perspective\, Nesterov's technique and many of its generaliz
 ations can be viewed as a systematic way to go from the continuous-time cu
 rves generated by the Bregman Lagrangian to a family of discrete-time acce
 lerated algorithms.\n	 \n
DTSTART:20161017T190000Z
DTEND:20161017T200000Z
LOCATION:Room 920\, Burnside Hall\, CA\, QC\, Montreal\, H3A 0B9\, 805 rue 
 Sherbrooke Ouest
SUMMARY:Andre Wibisono (University of Wisconsin-Madison)
URL:https://www.mcgill.ca/mathstat/channels/event/andre-wibisono-university
 -wisconsin-madison-263437
END:VEVENT
END:VCALENDAR