Significance tests using conditional distance covariance for the general formulation of asynchronous functional concurrent models

L. Freijeiro González, W. González Manteiga, M. Febrero Bande

Given two random vectors of arbitrary dimensions, X and Y, conditional independence testing allow one to verify if these are independent conditioned to a third one, Z. We briefly review this implication and introduce the novel measure of conditional dependence of Wang et al. (2015): the conditional distance covariance coefficient (CDC). A discussion about its high computational cost and strategies to make its estimation possible in practice follows. Eventually, we illustrate its utility in a particular framework: the asynchronous functional concurrent model (AFCM). Our proposal settles to employ the CDC term to develop new significance tests for the AFCM, providing a novel methodology. This resorts to nonparametric techniques considering the local character of the data to deal with its asynchronous nature.


Wang, X., Pan, W., Hu, W., Tian, Y., and Zhang, H. (2015). Conditional distance correlation. Journal of the American Statistical Association, 110(512):1726–1734.

Keywords: Asynchronous functional concurrent model (AFCM), Conditional distance covariance (CDC), Significance tests

GT09.NOPAR1 Invited Session. High dimension inference
November 7, 2023  3:30 PM
CC2: Conference Room

S. Novo Díaz, G. Aneiros Pérez

E. García Portugués, A. Meilán Vila

A. Lago, I. Van Keilegom, J. C. Pardo Fernández, J. de Uña Álvarez