Nicola Perra (Queen Mary University of London)

Talk Title: Past, present, and (possible) future of epidemic behavioural models
Abstract: Capturing the feedback loop between the transmission of infectious diseases and human behaviour is still a major challenge in Epidemiology. Most of the legacy solutions found in the literature can be classified as analytical behavioural feedback models. These are theoretical frameworks that incorporate non-linear mechanisms describing how individual behaviours change in response to the epidemic's progression. Most of the current frameworks instead, developed since the COVID-19 Pandemic, can be classified as data-driven behavioural feedback models. These integrate real-world data on behavioural changes, such as mobility patterns, social distancing measures, or other proxies into epidemic simulations. 
In this context, I will first present a systematic comparison of the performance of different behavioural feedback models considering the first wave of the COVID-19 pandemic, nine geographies and two modelling tasks. I will then introduce a novel modelling framework based on generalized contact matrices that aims to enhance epidemic models by capturing more realistically variations in behaviours among groups of the population.