LCN2 Seminar: Prediction of epidemics on networks
- Friday 27 November 2020
- Kaltura Live Room
Go to the Kaltura Live RoomEnter the Live Room
41st LCN2 seminar
Speaker: Bastian Prasse
Title: Prediction of epidemics on networks
During the epidemic outbreak of a virus, perhaps the greatest concern is the future evolution: how many people will be infected and which regions will be affected the most? The spread of a virus crucially depends on the contact network that specifies which individuals are in contact (e.g., closer than 1.5 meters). Hence, to predict an epidemic, it seems necessary to know the contact network, at least to some extent. But an accurate description of the contact network is a major challenge, due to the tremendous number of individuals and the heterogeneity of contact patterns, which range from isolated individuals to superspreaders.
My talk deals with the prediction of epidemics on unknown contact networks and consists of two parts. First, I introduce a prediction method, which uses past observations of the epidemic as input, infers the network as an intermediate step and returns predictions of the epidemic as output. Counterintuitively, the structure of the contact network is not necessary for an accurate prediction. Evaluations on data of the COVID-19 pandemic suggest that our prediction method outperforms several alternative approaches.
In the second part, I take a closer look at the interplay of the contact network and the epidemic outbreak. More specifically, we have derived the solution of the Susceptible-Infectious-Susceptible (SIS) epidemic model when the basic reproduction number R0 is close to one. Our derivations apply to arbitrarily large and heterogeneous contact networks. In agreement with the first part of the talk, our results show that greatly different networks result in the same epidemic outbreak, which is an important step towards a better understanding of epidemics on networks.
About the LCN2 seminar
This talk is part of a series of seminars organized within an ongoing scientific initiative called the "Leiden Complex Networks Network" (LCN2), which brings together scientists with a common interest in both theoretical models and empirical analyses of complex networks and random graphs. The LCN2 community shares the approach of using networks for describing real-world complex systems and aims at developing related analytical and numerical methods, while also being open to other research approaches for studying complex systems. The talks are designed for a broad audience, allowing for constructive exchanges of ideas between scientists from different disciplines. During and after the talk, some drinks and simple snacks are provided.