Inferring dynamics from data in rotavirus epidemiology

Abstract

The focus of this thesis is the inference of changes in the dynamics of rotavirus epidemiological data. As was discussed by S Hahné et al. [1], there was an exceptionally low rotavirus incidence in the Netherlands in the winter of 2013=2014. Motivated by an internal report from the National Institute of Public Health and the Environment (RIVM) [2] that provided a transmission model of rotavirus dynamics that suggested the appearance of bifurcations, we try to detect such bifurcations by analysing rotavirus time series with the use of Wasserstein distances (as is discussed by Michael Muskulus and Verduyn-Lunel in [3] for time series in general). Although we did not manage to detect the possible period doubling bifurcation affecting the Netherlands, we could use theWasserstein distances approach to detect changes in the dynamics of rotavirus corresponding to the introduction of vaccination against the disease in Germany.