Estimation of Functional Regression Models for Functional Responses by Wavelet Approximation
Ana Maria Aguilera, F. Ocaña, Mariano J. Valderrama · Contributions to statistics · 2008
A linear regression model to estimate a sample of response curves (realizations of a functional response) from a sample of predictor curves (functional predictor) is considered. Difierent procedures for estimating the parameter function of the model based on wavelets expansions and functional principal component decomposition of both the predictor and response curves are proposed. Wavelets coeficients will be estimated from discrete observations of sample curves at irregularly spaced time points that could be difierent among sample individuals.