References & Citations
Mathematics > Statistics Theory
Title: Learning extremal graphical structures in high dimensions
(Submitted on 1 Nov 2021 (v1), last revised 7 Jul 2022 (this version, v3))
Abstract: Extremal graphical models encode the conditional independence structure of multivariate extremes. For the popular class of H\"usler--Reiss models, we propose a majority voting algorithm for learning the underlying graph from data through $L^1$ regularized optimization. We derive explicit conditions that ensure consistent graph recovery for general connected graphs. A key statistic in our method is the empirical extremal variogram. We prove non-asymptotic concentration bounds for this quantity that hold for general multivariate Pareto distributions and are of independent interest.
Submission history
From: Michaël Lalancette [view email][v1] Mon, 1 Nov 2021 11:20:52 GMT (37kb)
[v2] Fri, 17 Jun 2022 04:54:04 GMT (364kb,D)
[v3] Thu, 7 Jul 2022 21:28:57 GMT (376kb,D)
Link back to: arXiv, form interface, contact.