When modeling the spread of an infection among members or nodes of a community, each node's probability of getting infected depends on its innate susceptibility and its exposure to the contagion through its neighbors.In many cases, a neighbor's influence is hidden. Such is the case with asymptomatic carriers of the disease. We develop generative models to identify the hidden influencers in both static and dynamic networks. We use the neighbors' hidden influence state to compute an accurate estimate of exposure to the contagion. We propose efficient variational inference algorithms to learn our models' parameters. We also study the causal mechanisms that lead to an elevated risk of infection.
If you would like to contact us about our work, please scroll down to the people section and click on one of the group leads' people pages, where you can reach out to them directly.