Abstract: For the degree corrected stochastic block model in the presence of arbitrary or even adversarial outliers, we develop a convex-optimization-based clustering algorithm that includes a penalization term depending on the positive deviation of a node from the expected number of edges to other inliers. We prove that under mild conditions, this method achieves exact recovery of the underlying clusters.

We test the performance of the algorithm on semi-synthetic heterogeneous networks reconstructed to match aggregate data on the Korean financial sector. Our method allows for recovery of sub-sectors with significantly lower error rates compared to existing algorithms. Our second application is to overlapping portfolio networks, for which we uncover a clustering structure.

Bio: Andreea Minca is an Associate Professor in the School of Operations Research and Information Engineering at Cornell University. She holds degrees from Sorbonne University (PhD in Applied Mathematics) and Ecole Polytechnique (Diplome de l'Ecole Polytechnique). Andreea received the 2016 SIAM Activity Group on Financial Mathematics and Engineering Early Career Prize, the NSF CAREER Award (2017), a Research Fellow of the Global Association of Risk Professionals (GARP) (2014), and an AXA Research Fund Awardee (2020). She serves on the editorial board of the SIAM Journal on Financial Mathematics.

Andreea Minca studies large systems under uncertainty, especially financial systems, and uses mathematical modeling to derive optimal policies that promote system stability. Her main work is on structural models for systemic risk, using networks to represent various types of interrelations. Her work has been published in leading Financial Mathematics and OR journals such as Mathematical Finance, Finance & Stochastics, SIAM Journal of Financial Mathematics, Operations Research and Management Science.