|
Abstract:
|
Nonparametric descriptions of network structure have flourished in recent years, driven by new computationally-feasible analysis methods and new classes of network models. Exchangeable networks are one such class that has taken root, with various proposed extensions including edge-exchangeability. This talk will give an overview of recent developments in this area, discussing the broad theoretical principles underpinning the modelling of stochastic arrays, and the inference of their structure. Various real data examples will demonstrate the power of this approach, and its applicability to repeated observations of networks.
|
