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: Distance-based Depths for Directional Data

Abstract: Directional data are constrained to lie on the unit sphere of~$\mathbb{R}^q$ for some~$q\geq 2$. To address the lack of a natural ordering for such data, depth functions have been defined on spheres. However, the depths available either lack flexibility or are so computationally expensive that they can only be used for very small dimensions~$q$. In this work, we improve on this by introducing a class of distance-based depths for directional data. Irrespective of the distance adopted, these depths can easily be computed in high dimensions too. We derive the main structural properties of the proposed depths and study how they depend on the distance used. We discuss the asymptotic and robustness properties of the corresponding deepest points. We show the practical relevance of the proposed depths in two applications, related to (i) spherical location estimation and (ii) supervised classification. For both problems, we show through simulation studies that distance-based depths have strong advantages over their competitors.
Comments: 19 pages, 5 figures
Subjects: Statistics Theory (math.ST)
Cite as: arXiv:1710.00080 [math.ST]
  (or arXiv:1710.00080v2 [math.ST] for this version)

Submission history

From: Davy Paindaveine [view email]
[v1] Fri, 29 Sep 2017 20:26:29 GMT (2477kb,D)
[v2] Thu, 2 Aug 2018 18:33:39 GMT (3369kb,D)

Link back to: arXiv, form interface, contact.