BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:&quot\;Improved ranking and selection for large-scale inference&qu
 ot\; - Prof Michael Newton\, University of Wisconsin-Madison
DTSTART:20160510T133000Z
DTEND:20160510T143000Z
UID:TALK66176@talks.cam.ac.uk
CONTACT:Alison Quenault
DESCRIPTION:Identifying leading measurement units from a large collection 
 is a common inference task in various domains of large-scale inference.\n\
 nTesting approaches\, which measure evidence against a null hypothesis rat
 her than effect magnitude\, tend to overpopulate lists of leading units wi
 th those associated with low measurement error. By contrast\, local maximu
 m likelihood (ML) approaches tend to favor units with high measurement err
 or.\n\nAvailable Bayesian and empirical Bayesian approaches rely on specia
 lized loss functions that result in similar deficiencies. We describe and 
 evaluate a generic empirical Bayesian ranking procedure that populates the
  list of top units in a way that maximizes the expected overlap between th
 e true and reported top lists for all list sizes. The procedure relates un
 it-specific posterior upper tail probabilities with their empirical distri
 bution to yield a ranking variable. It discounts high-variance units less 
 than popular non-ML methods and thus achieves improved operating character
 istics in the models considered.  Examples from genomics and sports are us
 ed to demonstrate the method.
LOCATION:Large  Seminar Room\, 1st Floor\, Institute of Public Health\, Un
 iversity Forvie Site\, Robinson Way\, Cambridge
END:VEVENT
END:VCALENDAR