B0457
Title: Approximate Bayesian inference in epidemic models: A focus on nowcasting and the time-varying reproduction number
Authors: Oswaldo Gressani - Hasselt University (Belgium) [presenting]
Abstract: In epidemiology, mathematical models play a determinant role in the analysis of infectious disease data. Statistical methods and their underlying algorithms form the core backbone to compute estimates of key epidemiological parameters and to quantify their associated uncertainty, thereby providing a robust toolbox to understand the disease dynamics resulting from the transmission of a pathogen in a population. When inference is carried out under the Bayesian paradigm, Markov chain Monte Carlo (MCMC) methods often require a large computational budget resulting in a prohibitively slow estimation process. Building upon the synergy between Laplace approximations and P-splines, a flexible methodology is proposed as a lightning-fast alternative to simulation-based methods. This new toolbox is illustrated in the context of nowcasting (i.e. the real-time assessment of the current epidemic situation corrected for imperfect data information caused by delays in reporting) and in the recently proposed EpiLPS framework for estimating the time-varying reproduction number with applications on data of SARS-CoV-2.