BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Sharp convergence guarantees for iterative (non)-convex empirical 
 risk minimization with random data - Dr. Kabir Verchand\, University of Ca
 mbridge
DTSTART:20231011T130000Z
DTEND:20231011T140000Z
UID:TALK206689@talks.cam.ac.uk
CONTACT:Dr Varun Jog
DESCRIPTION: Fitting a model to data typically involves applying an iterat
 ive algorithm to minimize an empirical risk.  However\, given a particular
  empirical risk minimization problem\, the process of algorithm selection 
 is often performed via either expensive trial-and-error or appeal to (pote
 ntially) conservative worst case efficiency estimates and it is unclear ho
 w to compare and contrast algorithms in a principled and meaningful manner
 .\nIn this talk\, we present one potential avenue to obtain fine-grained\,
  principled comparisons between iterative algorithms.  We provide a framew
 ork—based on Gaussian comparison inequalities—to characterize the traj
 ectory of an iterative algorithm run with sample-splitting on a set of non
 convex model-fitting problems with Gaussian data.  We use this framework t
 o demonstrate concrete separations in the convergence behavior of several 
 algorithms as well as to reveal some nonstandard convergence phenomena.
LOCATION:MR5\, CMS Pavilion A
END:VEVENT
END:VCALENDAR