General tests of conditional independence based on empirical processes indexed by functions - Inria - Institut national de recherche en sciences et technologies du numérique Accéder directement au contenu
Article Dans Une Revue Japanese Journal of Statistics and Data Science Année : 2023

General tests of conditional independence based on empirical processes indexed by functions

Résumé

This paper focuses on nonparametric procedures for testing conditional independence between random vectors using Möbius transformation. We derive a method predicated on general empirical processes indexed by a specific class of functions. Conditional half-space and conditional empirical characteristic processes are used to demonstrate two abstract approximation theorems and their applications in real-world situations. We conclude by describing the limiting behavior of the Möbius transformation of the empirical conditional processes indexed by functions under contiguous sequences of alternatives. Our results are proved under some standard structural conditions on the Vapnik-Chervonenkis classes of functions and some mild conditions on the model. Monte Carlo simulation results indicate that the suggested statistical test for indepen- dence behaves reasonably well in finite samples.
Fichier non déposé

Dates et versions

hal-04006333 , version 1 (27-02-2023)

Identifiants

Citer

Salim Bouzebda. General tests of conditional independence based on empirical processes indexed by functions. Japanese Journal of Statistics and Data Science , In press, ⟨10.1007/s42081-023-00193-3⟩. ⟨hal-04006333⟩
12 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More