On Azadkia-Chatterjee's conditional dependence coefficient

DOI10.3150/22-BEJ1529WikidataQ128905575 ScholiaQ128905575MaRDI QIDQ6201829FDOQ6201829

Authors: Hongjian Shi, Mathias Drton, Fang Han 

Publication date: 26 March 2024

Published in: Bernoulli (Search for Journal in Brave)

Abstract: In recent work, Azadkia and Chatterjee (2021) laid out an ingenious approach to defining consistent measures of conditional dependence. Their fully nonparametric approach forms statistics based on ranks and nearest neighbor graphs. The appealing nonparametric consistency of the resulting conditional dependence measure and the associated empirical conditional dependence coefficient has quickly prompted follow-up work that seeks to study its statistical efficiency. In this paper, we take up the framework of conditional randomization tests (CRT) for conditional independence and conduct a power analysis that considers two types of local alternatives, namely, parametric quadratic mean differentiable alternatives and nonparametric H"older smooth alternatives. Our local power analysis shows that conditional independence tests using the Azadkia--Chatterjee coefficient remain inefficient even when aided with the CRT framework, and serves as motivation to develop variants of the approach; cf. Lin and Han (2022b). As a byproduct, we resolve a conjecture of Azadkia and Chatterjee by proving central limit theorems for the considered conditional dependence coefficients, with explicit formulas for the asymptotic variances.

Full work available at URL: https://arxiv.org/abs/2108.06827

zbMATH Keywords

conditional independencerank-based testlocal power analysisgraph-based testnearest neighbor graphs

Mathematics Subject Classification ID

Statistics (62-XX) Probability theory and stochastic processes (60-XX)

Cites Work

• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• A General Asymptotic Framework for Distribution-Free Graph-Based Two-Sample Tests
• A NONPARAMETRIC HELLINGER METRIC TEST FOR CONDITIONAL INDEPENDENCE
• A Quality Index Based on Data Depth and Multivariate Rank Tests
• A consistent characteristic function-based test for conditional independence
• A copula-based non-parametric measure of regression dependence
• A multivariate two-sample test based on the number of nearest neighbor type coincidences
• A nearest neighbor estimate of the residual variance
• A new coefficient of correlation
• A simple measure of conditional dependence
• Asymptotic Statistics
• Conditional Distance Correlation
• Conditional Independence Testing in Hilbert Spaces with Applications to Functional Data Analysis
• Conditional independence for statistical operations
• Consistent nonparametric regression. Discussion
• Distribution and quantile functions, ranks and signs in dimension \(d\): a measure transportation approach
• Distribution-Free Consistent Independence Tests via Center-Outward Ranks and Signs
• Efficient multivariate entropy estimation via \(k\)-nearest neighbour distances
• Elements of causal inference. Foundations and learning algorithms
• Energy statistics: a class of statistics based on distances
• Estimation of a conditional copula and association measures
• Global sensitivity analysis: a novel generation of mighty estimators based on rank statistics
• Handbook of Graphical Models
• Introduction to nonparametric estimation
• Lectures on the nearest neighbor method
• Minimax optimal conditional independence testing
• Multivariate Rank-Based Distribution-Free Nonparametric Testing Using Measure Transportation
• Multivariate generalizations of the Wald-Wolfowitz and Smirnov two-sample tests
• Normal approximation under local dependence.
• On boosting the power of Chatterjee’s rank correlation
• On the fraction of random points by specified nearest-neighbour interrelations and degree of attraction
• On the multivariate runs test
• On the power of Chatterjee’s rank correlation
• On universally consistent and fully distribution-free rank tests of vector independence
• Panning for Gold: ‘Model-X’ Knockoffs for High Dimensional Controlled Variable Selection
• Partial distance correlation with methods for dissimilarities
• Sums of functions of nearest neighbor distances, moment bounds, limit theorems and a goodness of fit test
• Testing Statistical Hypotheses
• Testing conditional independence of discrete distributions
• Testing conditional independence using maximal nonlinear conditional correlation
• Testing conditional independence via Rosenblatt transforms
• Testing conditional independence via empirical likelihood
• The Hardness of Conditional Independence Testing and the Generalised Covariance Measure
• The Large-Sample Power of Tests Based on Permutations of Observations
• The asymptotic efficacies and relative efficiencies of various linear rank tests for independence
• The conditional permutation test for independence while controlling for confounders
• The expected size of some graphs in computational geometry

Cited In (1)

• Asymptotic expected sensitivity function and its applications to measures of monotone association

This page was built for publication: On Azadkia-Chatterjee's conditional dependence coefficient

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6201829)