Conditional mean and quantile dependence testing in high dimension (Q1747737)

The paper is concerned with the development of ``new tests for testing conditional mean and conditional quantile independence in the high-dimensional setting''. The tests are model-free and the authors claim that they capture nonlinear dependence. It appears as a nonparametric model-free alternative for the model developed in [\textit{P.-S. Zhong} and \textit{S. X. Chen}, J. Am. Stat. Assoc. 106, No. 493, 260--274 (2011; Zbl 1396.62110)]. Nevertheless, it is under linearity, less powerful. They sustain the claims by proving 6 theoretical results (propositions, theorems), and by examples which illustrate numerically the behavior of the proposals.