A useful parametric specification to model epidemiological data

Marco Mingione (Roma Tre University)

Monday 20th March 13:00-14:00 Maths 311B


Recent epidemic outbreaks reignited the debate on how to properly model epidemiological indicators in order to (i) gain insights into the dynamics of infectious diseases, (ii) generate epidemic forecasts and  (iii) implement timely prevention policies. We argue that the five-parameter specification of the Richards' curve, often used in the context of biological growth processes, can address most of the features occurring in epidemiological data. We propose an extension to the standard parametric specification which envisions the presence of an endemic rate, accounting for the possibility that future epidemics will not be eradicated. Two different estimation methods are described: the first, based on likelihood maximization, is particularly useful when the outbreak is still ongoing and the main goal is to obtain sufficiently accurate estimates in negligible computational run-time. The second is fully Bayesian, and allows for more ambitious modelling attempts such as the inclusion of spatial and temporal dependence, but it requires more data and computational resources. To demonstrate such flexibility, we show different applications using publicly available data on recent epidemics where the data collection processes and transmission patterns are extremely heterogeneous. Results highlight good fitting performances and reasonable short - to medium-term forecasts.

Add to your calendar

Download event information as iCalendar file (only this event)