COSTNET: the European Cooperation for Statistics of Network Data Science.
This page provides an overview of the planned Working Groups with a preliminary work programme.
Chairs: Mariaclelia Di Serio and Silvia Fierascu.
This group will deal with sampling data from, of and on networks. How can one determine whether a viral advertising campaign was successful? What is the difference between sampling a gene regulatory network in a transsectional or longitudinal manner and does it matter? How can one sample from highly or poorly connected nodes in an infectious disease network?
Chairs: Steffen Lauritzen and Konstantin Avrachenkov
Percolation models, diffusion models, graphical models, ordinary and stochastic differential equation models and many other models have been proposed to describe networks and their behaviour. The WG will focus on comparing and developing network models for existing and novel applications, such as finance, sociology, epidemiology and biology.
Chairs: Arnoldo Frigessi and Nial Friel
Network inference consists of computationally identifying network model parameters from data and as such it builds forth on the activities of the first two groups. Inference provides explanations and descriptions of phenomena. Network prediction builds forth on inference to deal with diverse questions, such as “What is the best advertising or vaccination strategy?”, “Is this network activity a sign of fraud?” and “Will this drug be effective?”.