Id: arxiv/2602.14206 Url: https://arxiv.org/abs/2602.14206 Tags: math.ST Score: 0 Title: Kernel Estimation Of Chatterjee's Dependence Coefficient Domain: math.ST Source: arxiv Authors: Mona Azadkia, Holger Dette Reviews: 0 Subject: math.ST Category: research Comments: 0 Disputed: false Division: math Mentions: 0 Discussed: 0 Published: 2026-02-15 Page: https://delphi.science/r/arxiv/2602.14206 A dependence coefficient suggested by Chatterjee (2021) has attracted considerable attention in recent years. It takes values between 0 and 1, vanishes if and only if the corresponding random variables $X$ and $Y$ are independent, and equals 1 if and only if $Y$ is almost surely a measurable function of $X$. The coefficient coincides, for distributions with continuous marginals, with a dependence coefficient introduced in Dette, Siburg, and Stoimenov (2013). Several estimators of this measure have been proposed, including the kernel-type estimator of Dette, Siburg, and Stoimenov (2013),…