BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Sharp oracle inequalities for stationary points of nonconvex penal
 ised M-estimators - Andreas Elsener (ETH Zürich)
DTSTART:20180118T144500Z
DTEND:20180118T153000Z
UID:TALK97804@talks.cam.ac.uk
CONTACT:INI IT
DESCRIPTION:<span>Co-author: Sara van de Geer		(ETH Zurich)        <br></s
 pan><span><br>Nonconvex loss functions are used in several areas of statis
 tics and machine learning. They have several appealing properties as in th
 e case of robust regression. We propose a general framework to derive shar
 p oracle inequalities for stationary points of penalised M-estimators with
  nonconvex loss. The penalisation term is assumed to be a weakly decomposa
 ble norm. We apply the general framework to sparse (additive) corrected li
 near regression\, sparse PCA\, and sparse robust regression. Finally\, a n
 ew estimator called "robust SLOPE" is proposed to illustrate how to apply 
 our framework to norms different from the l1-norm.&nbsp\;</span>
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR