|
Abstract:
|
Individual contact networks are rarely available concurrently with an epidemic. Using empirical likelihood methodology, we propose a robust method to analyze infectious disease spread over an unknown contact network within the SEIR framework. The method allows trivially parallelizable approximate Bayesian computation methods to be employed for model fitting, which, in turn, allows for up-to-the-day analyses of current epidemics. The model is tested via a new analysis of the Abakaliki smallpox epidemic of 1967.
|
