We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

math.ST

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Mathematics > Statistics Theory

Title: Graphical models for infinite measures with applications to extremes and Lévy processes

Abstract: Conditional independence and graphical models are well studied for probability distributions on product spaces. We propose a new notion of conditional independence for any measure $\Lambda$ on the punctured Euclidean space $\mathbb R^d\setminus \{0\}$ that explodes at the origin. The importance of such measures stems from their connection to infinitely divisible and max-infinitely divisible distributions, where they appear as L\'evy measures and exponent measures, respectively. We characterize independence and conditional independence for $\Lambda$ in various ways through kernels and factorization of a modified density, including a Hammersley-Clifford type theorem for undirected graphical models. As opposed to the classical conditional independence, our notion is intimately connected to the support of the measure $\Lambda$. Our general theory unifies and extends recent approaches to graphical modeling in the fields of extreme value analysis and L\'evy processes. Our results for the corresponding undirected and directed graphical models lay the foundation for new statistical methodology in these areas.
Comments: 48 pages, 4 figures, 1 table
Subjects: Statistics Theory (math.ST); Probability (math.PR); Methodology (stat.ME)
MSC classes: 62H22, 60E07, 60G70, 60G51
Cite as: arXiv:2211.15769 [math.ST]
  (or arXiv:2211.15769v1 [math.ST] for this version)

Submission history

From: Kirstin Strokorb [view email]
[v1] Mon, 28 Nov 2022 20:58:48 GMT (98kb,D)

Link back to: arXiv, form interface, contact.