Title: Structure in complex networks
Abstract: A core challenge in data science is obtaining a principled understanding of the structure of empirical data. The particular case of complex network data is increasingly important and particularly challenging due, for example, to dependencies among samples and lack of obvious spatial structure. I will discuss recent advances on this problem, including new data models, approximation theorems, scalable algorithms, and applications. These advances apply to data with a number of different structures, such as community structure, core-periphery structure, temporality, and multiplexity. The analytic and algorithmic tools are drawn from a variety of fields, such as metallurgy, compressed sensing, and statistics. Applications include image segmentation, counterterrorism, neuroscience, and genealogy.