BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Cox process representation and inference for stochastic reaction-d
 iffusion processes - David Schnoerr (University of Edinburgh)
DTSTART:20160608T140000Z
DTEND:20160608T150000Z
UID:TALK66392@talks.cam.ac.uk
CONTACT:INI IT
DESCRIPTION:Complex behaviour in many systems arises from the stochastic i
 nteractions of spatially distributed particles or agents. Stochastic react
 ion-diffusion processes are widely used to model such behaviour in discipl
 ines ranging from systems biology to the social sciences\, yet they are no
 toriously difficult to simulate and calibrate to observational data. On th
 e other hand\, spatio-temporal point processes offer several computational
  advantages from the statistical perspective. In this talk\, I will show h
 ow the Poisson representation of the Chemical Master Equation can be used 
 to derive a novel connection between stochastic reaction-diffusion process
 es and spatio-temporal Cox processes. This connection allows us to natural
 ly define an approximate likelihood\, which can be optimised with respect 
 to the kinetic parameters of the model. We show on several examples from s
 ystems biology and epidemiology that the method yields consistently accura
 te parameter estimates\, and can be used effectively for model selection.
LOCATION:Seminar Room 2\, Newton Institute
END:VEVENT
END:VCALENDAR