School on Nonlinear Dynamics, Complex Networks, Information Theory and Machine Learning in Neuroscience

Epidemiological models, like meteorological models, have their origins in simple equations that nevertheless managed to capture the essence of the phenomena studied. Two physicians, Ronald Ross and Anderson G. McKendrick (both closely related to India), and a chemist, William O. Kermack, laid the foundations of epidemiological modeling around the same time that physicist Lewis Fry Richardson proposed the first weather prediction models. Despite their success in explaining the phenomenology observed in contagion processes, the practical application of epidemiological models was not comparable to those dedicated to meteorology, at least in terms of their predictive nature. The arrival of the 21st century brought with it the democratization of the Internet which, together with advances in the capture of data on human activity, made it possible to describe the skeleton of interactions through which infectious diseases are transmitted. In this colloquium we will describe the efforts invested during the last two decades to incorporate into epidemic models ingredients such as the complexity of contact networks, patterns of human mobility at different geographical scales and the different temporal scales associated with human interactions. With this information available, the theoretical and computational tools of complex systems and network science have demonstrated their usefulness in analyzing real epidemiological problems, developing theoretical and computational frameworks that allow not only to accurately reproduce the spatiotemporal trajectory of an epidemic, but also to evaluate epidemiological control strategies.