Activity Number:
|
69
- Network Analysis
|
Type:
|
Contributed
|
Date/Time:
|
Monday, August 3, 2020 : 10:00 AM to 2:00 PM
|
Sponsor:
|
Section on Statistical Learning and Data Science
|
Abstract #313757
|
|
Title:
|
Anomaly Detection on Complex Networks via Topological Data Analysis
|
Author(s):
|
Ignacio Segovia-Dominguez* and Dorcas Ofori-Boateng and Cuneyt Gurcan Akcora and Murat Kantarcioglu and Yulia Gel
|
Companies:
|
The University of Texas at Dallas and NASA Jet Propulsion Laboratory and University of Manitoba and The University of Texas at Dallas and University of Texas at Dallas
|
Keywords:
|
anomaly detection;
dynamic networks;
topological data analysis;
blockchain
|
Abstract:
|
We introduce a new nonparametric method for anomaly detection on graphs using the emerging tools of topological data analysis. Our method is based on accounting for changes in a topological and geometric structure of dynamic networks under the framework of persistent homology. In particular, we describe shapes of complex networks via analysis of clique communities and then track fluctuations of the resulting topological signatures over time. We validate the proposed approach in application to unilayer and multilayer networks from social sciences, telecommunication and blockchain.
|
Authors who are presenting talks have a * after their name.