Saved in:
Bibliographic Details
Main Authors: He, Shuaida, Chen, Yangzhou, Chen, Xin
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2604.10482
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915957530689536
author He, Shuaida
Chen, Yangzhou
Chen, Xin
author_facet He, Shuaida
Chen, Yangzhou
Chen, Xin
contents Modern regression analysis often involves responses and predictors taking values in the same or distinct metric spaces. To rank non-Euclidean heterogeneous predictors in regression by explanatory strength, analogous to the classical $R^2$, we introduce the Fréchet correlation coefficient (FCC), defined as the relative reduction in the Fréchet variance of the response after conditioning on a specific predictor. FCC is directional, model-free, and interpretable on a unit-scale, attaining one under almost sure functional dependence and zero when the Fréchet mean is invariant to conditioning. We propose a novel partition-based estimator that avoids explicit nonparametric estimation of the conditional Fréchet mean function, thereby improving both computational efficiency and flexibility. A tailored wild bootstrap algorithm is further developed for testing the Fréchet conditional mean dependence. We establish asymptotic theory and evaluate power through extensive simulations.
format Preprint
id arxiv_https___arxiv_org_abs_2604_10482
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle The Fréchet correlation coefficient for heterogeneous random objects
He, Shuaida
Chen, Yangzhou
Chen, Xin
Methodology
Modern regression analysis often involves responses and predictors taking values in the same or distinct metric spaces. To rank non-Euclidean heterogeneous predictors in regression by explanatory strength, analogous to the classical $R^2$, we introduce the Fréchet correlation coefficient (FCC), defined as the relative reduction in the Fréchet variance of the response after conditioning on a specific predictor. FCC is directional, model-free, and interpretable on a unit-scale, attaining one under almost sure functional dependence and zero when the Fréchet mean is invariant to conditioning. We propose a novel partition-based estimator that avoids explicit nonparametric estimation of the conditional Fréchet mean function, thereby improving both computational efficiency and flexibility. A tailored wild bootstrap algorithm is further developed for testing the Fréchet conditional mean dependence. We establish asymptotic theory and evaluate power through extensive simulations.
title The Fréchet correlation coefficient for heterogeneous random objects
topic Methodology
url https://arxiv.org/abs/2604.10482