Translator Disclaimer
October 2019 Distance multivariance: New dependence measures for random vectors
Björn Böttcher, Martin Keller-Ressel, René L. Schilling
Ann. Statist. 47(5): 2757-2789 (October 2019). DOI: 10.1214/18-AOS1764

Abstract

We introduce two new measures for the dependence of $n\ge2$ random variables: distance multivariance and total distance multivariance. Both measures are based on the weighted $L^{2}$-distance of quantities related to the characteristic functions of the underlying random variables. These extend distance covariance (introduced by Székely, Rizzo and Bakirov) from pairs of random variables to $n$-tuplets of random variables. We show that total distance multivariance can be used to detect the independence of $n$ random variables and has a simple finite-sample representation in terms of distance matrices of the sample points, where distance is measured by a continuous negative definite function. Under some mild moment conditions, this leads to a test for independence of multiple random vectors which is consistent against all alternatives.

Citation

Download Citation

Björn Böttcher. Martin Keller-Ressel. René L. Schilling. "Distance multivariance: New dependence measures for random vectors." Ann. Statist. 47 (5) 2757 - 2789, October 2019. https://doi.org/10.1214/18-AOS1764

Information

Received: 1 December 2017; Revised: 1 September 2018; Published: October 2019
First available in Project Euclid: 3 August 2019

zbMATH: 07114928
MathSciNet: MR3988772
Digital Object Identifier: 10.1214/18-AOS1764

Subjects:
Primary: 62H20
Secondary: 60E10, 62G10, 62G15, 62G20

Rights: Copyright © 2019 Institute of Mathematical Statistics

JOURNAL ARTICLE
33 PAGES


SHARE
Vol.47 • No. 5 • October 2019
Back to Top