Mathmod 2018 Extended Abstracts

Towards Bayesian Parametrization of National Scale Epidemics

MATHMOD 2018 Extended Abstract Volume​, ARGESIM Report 55 (ISBN 978-3-901608-91-9), p 65-66, DOI: 10.11128/arep.55.a55245


Many infections or diseases that pose a public health threat have a zoonotic origin. Verotoxin-producing Escherichia coli (VTEC) is an example of a zoonotic foodborne pathogen where cattle can act as a reservoir. We simulate the spreading of VTEC on the Swedish cattle network using transportation data and a compartment model in the software SimInf. With the model and available data, we perform Bayesian inference using simulation-based likelihood-free methods, namely Approximate Bayesian Computations and Synthetic likelihood. By using our approach, we show that epidemical models on a national scale can be parameterized using observed data.