A1211
Title: A kernel-based goodness-of-fit test for functional regression models
Authors: Daniel Diz-Castro - Centro de investigacion tecnoloxia matematica de galicia (CITMAGA) (Spain) [presenting]
Manuel Febrero-Bande - University of Santiago de Compostela (Spain)
Wenceslao Gonzalez-Manteiga - University of Santiago de Compostela (Spain)
Abstract: The problem of assessing the goodness-of-fit (GoF) of parametric regression models with functional covariates is addressed. A robust kernel-based methodology is presented, and theoretical results previously obtained in the Euclidean setting are shown to extend to more general frameworks in which both the covariates and the parameters are infinite-dimensional. A computationally tractable test statistic is introduced, and its asymptotic properties are analysed. For calibration purposes, a consistent multiplier bootstrap procedure is proposed. The finite-sample performance of the method is illustrated through both a simulation study and a real data application.
