BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Strong converses and high-dimensional statistical estimation probl
 ems - Ramji Venkataramanan (University of Cambridge)
DTSTART:20180724T084500Z
DTEND:20180724T093000Z
UID:TALK108283@talks.cam.ac.uk
CONTACT:INI IT
DESCRIPTION:<span>In many statistical inference problems\, we wish to boun
 d the performance of any possible estimator. This can be seen as a convers
 e result\, in a standard information-theoretic sense. A standard approach 
 in the statistical literature is based on Fano&rsquo\;s inequality\, which
  typically gives a weak converse. We adapt these arguments by replacing Fa
 no by more recent information-theoretic ideas\, based on the work of Polya
 nskiy\, Poor and Verdu. This gives tighter lower bounds that can be easily
  computed and are asymptotically sharp. We illustrate our technique in thr
 ee applications: density estimation\, active learning of a binary classifi
 er\, and compressed sensing\, obtaining tighter risk lower bounds in each 
 case. &nbsp\; <br> <br> (joint with Oliver Johnson\, see doi:10.1214/18-EJ
 S14)</span>  <br><br><br><br><br>
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR