BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Which survival model? Model selection and model averaging issues f
 or analysis of survival and event history data - Prof. Nils Hjort\, Depart
 ment of Mathmatics\, University of Oslo\, Norway
DTSTART:20120117T143000Z
DTEND:20120117T153000Z
UID:TALK34826@talks.cam.ac.uk
CONTACT:Li Su
DESCRIPTION:Most model selection methods work in "overall modus"\, as for\
 nthe AIC\, BIC\, etc.\; one is then expected to you use what one gets\,\nr
 egardless of the actual purpose of the analysis. Here I focus the\nmachine
 ry\, looking for models that work well for a given purpose\,\nand in parti
 cular choose to accept and then work with the view\nthat even with the sam
 e data set and the same list of candidate\nmodels\, one model may be best 
 for explaining e.g. survival for\nmen above sixty whereas another model ma
 y be better for\ne.g. estimating median remaining survival time for women\
 , etc.\nThis ideas in particular lead to version of FIC (focussed informat
 ion\ncriteria)\, for multiplicative Cox regression and for additive Aalen\
 nregression.
LOCATION:Large  Seminar Room\, 1st Floor\, Institute of Public Health\, Un
 iversity Forvie Site\, Robinson Way\, Cambridge
END:VEVENT
END:VCALENDAR