On boosting the power of Chatterjee’s rank correlation

DOI10.1093/BIOMET/ASAC048zbMATH Open1516.62059arXiv2108.06828OpenAlexW3195538177MaRDI QIDQ6045142FDOQ6045142

Authors: Fang Han 

Publication date: 26 May 2023

Published in: Biometrika (Search for Journal in Brave)

Abstract: Chatterjee (2021)'s ingenious approach to estimating a measure of dependence first proposed by Dette et al. (2013) based on simple rank statistics has quickly caught attention. This measure of dependence has the unusual property of being between 0 and 1, and being 0 or 1 if and only if the corresponding pair of random variables is independent or one is a measurable function of the other almost surely. However, more recent studies (Cao and Bickel, 2020; Shi et al., 2021b) showed that independence tests based on Chatterjee's rank correlation are unfortunately rate-inefficient against various local alternatives and they call for variants. We answer this call by proposing revised Chatterjee's rank correlations that still consistently estimate the same dependence measure but provably achieve near-parametric efficiency in testing against Gaussian rotation alternatives. This is possible via incorporating many right nearest neighbors in constructing the correlation coefficients. We thus overcome the "only one disadvantage" of Chatterjee's rank correlation (Chatterjee, 2021, Section 7).

Full work available at URL: https://arxiv.org/abs/2108.06828

Recommendations

• On the asymptotic null distribution of the symmetrized Chatterjee's correlation coefficient
• Data-Driven Rank Tests for Independence
• A new rank correlation measure
• Measuring and testing dependence by correlation of distances
• High-dimensional consistent independence testing with maxima of rank correlations

zbMATH Keywords

independence testrank correlationdependence measurelocal power analysisright nearest neighbour

Mathematics Subject Classification ID

Hypothesis testing in multivariate analysis (62H15) Measures of association (correlation, canonical correlation, etc.) (62H20)

Cited In (9)

• On the asymptotic null distribution of the symmetrized Chatterjee's correlation coefficient
• Nonlinear directed acyclic graph estimation based on the kernel partial correlation coefficient
• Combining dissimilarity matrices by using rank correlations
• Asymptotic expected sensitivity function and its applications to measures of monotone association
• Distribution-free tests of multivariate independence based on center-outward quadrant, Spearman, Kendall, and van der Waerden statistics
• On Azadkia-Chatterjee's conditional dependence coefficient
• Exact detection thresholds and minimax optimality of Chatterjee's correlation coefficient
• Rearranged dependence measures
• A Kernel Measure of Dissimilarity between M Distributions

This page was built for publication: On boosting the power of Chatterjee’s rank correlation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6045142)