Disease spread can be modelled by compartmental models, such as the SEIR (Susceptible, Exposed, Infected, Recovered) model, that are described by a deterministic dynamical system usually defined in terms of a system of ordinary differential equations. Even though they are simplifications of the real disease spread dynamics, they can be quite useful for assessing the stage of a disease and forecasting associated risks. However, their calibration typically relies on low-quality and sparse data, and classical calibration methods do not scale well for more complex, spatial models. Nevertheless, recent advances in machine learning methods have made it possible to properly calibrate more complex models, in particular those that take into account human mobility and population dynamics. As evidenced in the COVID-19 pandemic, human mobility can play a significant role in disease spread and developing mathematical models that consider the effects of it is an important research endeavour.

Projects will involve developing and studying the properties of compartmental epidemiological models coupled with human mobility and population dynamics, as well as developing, implementing and applying calibration methods based on machine learning techniques, such as neural networks. Models can be developed based on real or hypothetical applications, and calibrated considering real or synthetic data.

References

Peixoto P. S.; Marcondes, D.; Peixoto, C.; Oliva S. M. Modeling future spread of infections via mobile geolocation data and population dynamics. An application to COVID-19 in Brazil. PLoS ONE 15(7): e0235732. 2020