A0906
Title: Kernel estimation in functional-circular regression models
Authors: Andrea Meilan-Vila - Universidade da Coruna (Spain) [presenting]
Rosa Crujeiras - University of Santiago de Compostela (Spain)
Mario Francisco-Fernandez - Universidade da Coruna (Spain)
Abstract: The aim is to consider a regression model featuring a circular response and a functional covariate. Within this setting, a Nadaraya-Watson-type estimator is proposed and explored. Its asymptotic bias, variance, and distribution are derived, thereby providing a comprehensive theoretical foundation. Furthermore, cross-validation methods are developed and assessed for bandwidth selection tailored to this framework. To evaluate the estimators' performance, a simulation study is conducted and its practical utility is illustrated using a real dataset consisting of daily temperature curves.
