Modeling the new flu wave using data science and complex networks theory.

This Master’s thesis explores the use of the SEINR (Susceptible, Exposed, Infectious, Non-infectious, Recovered) compartmental model to forecast the evolution of flu-like ill- ness cases in Italy during the 2023-24 winter season. The model’s innovative aspects lie in its region-based meta-population framework, which simulates intra- and inter-regional mobility, capturing network dynamics critical to understanding how a disease spreads in Italy’s diverse demographic and geographic landscape. This approach, previously successful in evaluating the efficacy of NPIs during the COVID-19 pandemic, is further refined by incorporating features such as class divisions based on age, activity levels, and vulnerability of different age groups. Key epidemiological parameters, such as infection duration and transmission rates, are then estimated with Bayesian inference methods to improve forecast accuracy. Empirical results demonstrate the model's effectiveness in capturing flu trends across Italian regions, especially in critical timeframes during which an increased number of contacts is observable. This work emphasizes the role of adaptive modeling in epidemiology, and how public health strategies driven by past data can help in managing seasonal epidemics.