Low-rank models in multiple-network analysis

Organizers: Keith Levin, Vince Lyzinski and Avanti Athreya

Description: In many domains of interest, most notably neuroimaging, data comes in the form of multiple networks that share some underlying structure. As a result, there is growing interest in problems related to multiple-network analysis and inference. Many of the initial statistical approaches to these problems have adapted or extended popular low-rank network models, such as the stochastic blockmodel and its generalizations. The result has been a flurry of new models for collections of networks, drawing on multilinear algebra, tensor methods and related areas. The aim of this workshop is

• to take stock of the work done so far in this direction, and
• to begin to lay out next steps and unifying aims.

While the focus of the workshop is statistical in nature, it will undoubtedly be of interest to researchers in computer science, engineering and physics, as well.