BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:BSU Seminar: &quot\;Upgrading survival models with CARE&quot\; - W
 illiam Underwood\, Statistical Laboratory\, University of Cambridge
DTSTART:20250923T130000Z
DTEND:20250923T140000Z
UID:TALK235120@talks.cam.ac.uk
CONTACT:Alison Quenault
DESCRIPTION:Clinical risk prediction models are regularly updated as new d
 ata\, often with additional covariates\, become available. We propose CARE
  (Convex Aggregation of relative Risk Estimators) as a general approach fo
 r combining existing "external" estimators\n with a new data set in a time
 -to-event survival analysis setting. Our method initially employs the new 
 data to fit a flexible family of reproducing kernel estimators via penalis
 ed partial likelihood maximisation. The final relative risk estimator is t
 hen constructed\n as a convex combination of the kernel and external estim
 ators\, with the convex combination coefficients and regularisation parame
 ters selected using cross-validation. We establish high-probability bounds
  for the L2-error of our proposed aggregated estimator\,\n showing that it
  achieves a rate of convergence that is at least as good as both the optim
 al kernel estimator and the best external model. Empirical results from si
 mulation studies align with the theoretical results\, and we illustrate th
 e improvements our methods\n provide for cardiovascular disease risk model
 ling.
LOCATION:Large Downstairs Teaching Room\, East Forvie Building\, Forvie Si
 te Robinson Way Cambridge CB2 0SR.
END:VEVENT
END:VCALENDAR