BEGIN:VCALENDAR
PRODID:-//AT Content Types//AT Event//EN
VERSION:2.0
METHOD:PUBLISH
BEGIN:VEVENT
DTSTAMP:20240620T155219Z
CREATED:20180920T091857Z
UID:ATEvent-dd5d358804e94b0c8323818f5cb20614
LAST-MODIFIED:20180924T200312Z
SUMMARY:Talk by Nicolas Keriven (ENS\, CFM-ENS Chair\, Paris) : A Dual Certificates Analysis of Compressive Off-the-Grid Recovery
DTSTART:20181019T080000Z
DTEND:20181019T090000Z
DESCRIPTION:Many problems in machine learning and imaging can be frame
d as an infinite dimensional Lasso problem to estimate a sparse measur
e. This includes for instance regression using a continuously paramete
rized dictionary\, mixture model estimation and super-resolution of im
ages. To make the problem tractable\, one typically sketches the obser
vations (often called compressive-sensing in imaging) using randomized
projections. In this work\, we provide a comprehensive treatment of t
he recovery performances of this class of approaches\, proving that (u
p to log factors) a number of sketches proportional to the sparsity is
enough to identify the sought after measure with robustness to noise.
We prove both exact support stability (the number of recovered atoms
matches that of the measure of interest) and approximate stability (lo
calization of the atoms) by extending two classical proof techniques (
minimal norm dual certificate and golfing scheme certificate).
LOCATION:Room M7.101
CLASS:PUBLIC
END:VEVENT
END:VCALENDAR