The Hellinger Correlation

10/24/2018
by   Gery Geenens, et al.
0

In this paper, the defining properties of a valid measure of the dependence between two random variables are reviewed and complemented with two original ones, shown to be more fundamental than other usual postulates. While other popular choices are proved to violate some of these requirements, a class of dependence measures satisfying all of them is identified. One particular measure, that we call the Hellinger correlation, appears as a natural choice within that class due to both its theoretical and intuitive appeal. A simple and efficient nonparametric estimator for that quantity is proposed. Synthetic and real-data examples finally illustrate the descriptive ability of the measure, which can also be used as test statistic for exact independence testing.

READ FULL TEXT
research
01/10/2022

Rearranged dependence measures

Most of the popular dependence measures for two random variables X and Y...
research
12/03/2022

A simple extension of Azadkia & Chatterjee's rank correlation to a vector of endogenous variables

In the present paper, we propose a direct and simple extension of Azadki...
research
12/03/2020

Modeling Count Data via Copulas

Copula models have been widely used to model the dependence between cont...
research
04/09/2023

Nearest-Neighbor Sampling Based Conditional Independence Testing

The conditional randomization test (CRT) was recently proposed to test w...
research
09/09/2020

The Earth Mover's Correlation

Since Pearson's correlation was introduced at the end of the 19th centur...
research
05/09/2021

Bayesian Kernelised Test of (In)dependence with Mixed-type Variables

A fundamental task in AI is to assess (in)dependence between mixed-type ...
research
05/09/2015

Measuring dependence powerfully and equitably

Given a high-dimensional data set we often wish to find the strongest re...

Please sign up or login with your details

Forgot password? Click here to reset