BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:What is the deviance information criterion? - Martyn Plummer\, Int
 ernational Agency for Research on Cancer
DTSTART:20071025T130000Z
DTEND:20071025T140000Z
UID:TALK8748@talks.cam.ac.uk
CONTACT:Nikolaos Demiris
DESCRIPTION:The deviance information criterion (DIC) is a Bayesian analogu
 e of the Akaike Information Criterion (AIC). Although DIC is widely used a
 s a model choice tool in Bayesian data analysis\, its theoretical foundati
 ons\nremain a subject of controversy.  I will describe a theoretical deriv
 ation of DIC as a penalized loss function\, where the penalty represents a
  rational price that must be paid for using the data twice:\nonce for para
 meter estimation\, and again for model criticism.  This derivation suggest
 s some limitations on situations where DIC may be used\, as well as provid
 ing improved asymptotic approximations and useful extensions.
LOCATION:Small Seminar Room\, 1st Floor\, Institute of Public Health\, Uni
 versity Forvie Site\, Robinson Way\, Cambridge
END:VEVENT
END:VCALENDAR