BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Online Learning and Online Convex Optimisation - Siddharth Swaroop
  (University of Cambridge)
DTSTART:20210414T100000Z
DTEND:20210414T113000Z
UID:TALK157390@talks.cam.ac.uk
CONTACT:Elre Oldewage
DESCRIPTION:In online learning\, data arrives sequentially\, and model par
 ameters are\nupdated at each step. This is in contrast to batch training\,
  where all\ndata is available at once. Recently\, online learning has larg
 e-scale\napplications such as online web ranking and online advertisement\
 nplacement\, and is closely related to continual learning. The field of\no
 nline learning itself is well-established\, with a lot of theory.\n\nWe wi
 ll closely follow "Online Learning and Online Convex Optimisation"\nby Sha
 lev-Shwartz (2011) (up to and including Section 2.5). We will see\nhow imp
 ortant convexity is\, and analyse the regret of some well-known\nalgorithm
 s such as Follow-The-Leader\, Follow-The-Regularised-Leader\, and\nOnline 
 Gradient Descent.
LOCATION:https://eng-cam.zoom.us/j/82019956685?pwd=WUNSVVcrdC9IZGxQOHFhSTh
 jUjd2dz09
END:VEVENT
END:VCALENDAR