BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Approximating Data - Laurie Davies\, Univeristy of Duisburg-Essen
DTSTART:20130426T150000Z
DTEND:20130426T160000Z
UID:TALK44680@talks.cam.ac.uk
CONTACT:Richard Samworth
DESCRIPTION:The talk will describe an approach to much of statistics in wh
 ich\nprobability models are consistently treated as approximations to the 
 data. It\nis not assumed that the data are distributed as in the model\, n
 or does\none behave as if this were true whilst being conscious of the fac
 t\nthat it is not.  A model P is regarded as an adequate approximation to\
 nthe data x of size n if 'typical' samples X(P) of size n simulated under 
 the\ndata 'look like' x. The words 'typical' and 'looks like' must be give
 n\nprecise meanings which will depend on the problem. The approach has\nse
 veral consequences some of which may be unexpected: there are no\n'true bu
 t unknown' parameter values and the interpretation of\nconfidence or appro
 ximation intervals is non-frequentist. Examples will be\ngiven ranging fro
 m the location-scale problem to non-parametric\nregression and image analy
 sis.
LOCATION:MR12\, CMS\, Wilberforce Road\, Cambridge\, CB3 0WB
END:VEVENT
END:VCALENDAR