Partial Likelihood for Space-Time Models
Partial likelihood methods allow spatio-temporal models to be fitted to point data without the need for computationally demanding Monte Carlo methods. They can be applied to any model that specifies the intensity of future events conditional on the pattern of past events. This occurs naturally in such diverse areas as infectious disease models or patterns of birds choosing nest locations on a finite habitat.
We are currently exploring ways to use partial likelihood methods to estimate important structural parameters of infectious disease outbreaks, for example a spatio-temporal extension of the reproduction number, R0.