Luogo: l'evento si terrà per via telematica attraverso collegamento alla piattaforma Zoom. Per richiedere le credenziali di accesso contattare la segreteria (dept.math [at] unitn.it).
- Nico Stollenwerk (Universidade de Lisboa, Portugal)
We investigate epidemiological models with susceptibles S, infected I and recovered R populations, where infected can be distinguished into severe and potentially hospitalized cases H, which are easily recordable, and mild or even asymptomatic cases A, which eventually could be detected but often stay unrecorded and still contribute to the force of infection, in so called SHAR models. Such SHAR models can be compared in likelihood functions and in a Bayesian framework with effective SIR models, where shifts in parameters are detected, and the concept of effective parameters varying from one model to the next is established. Applications will be given from dengue fever epidemiology to corona virus disease 2019, COVID-19, and model refinements uppon data information discussed.
Referente: Andrea Pugliese