Title: From Motifs to Persistent Homologies: What the Network Geometry Tells Us on Functionality of Complex Networks

Abstract: Currently available methods for complex network analysis largely focus on global characteristics and do not account for local higher order network properties of graph-structured data. However, a number of recent studies indicate that the local intrinsic network geometry may be the key for our understanding the hidden mechanisms behind organization of complex network systems.  In this talk we discuss new data science methods, based on a combination of topological data analysis on graphs and network motif inference, for analysis of a functional role of persistent higher-order network features in characterizing formation, robustness and resiliency of complex networks, with applications to power grids and Bitcoin.

Professor Yulia Gel's webpage is http://www.utdallas.edu/~yxg142030/