EcoSta 2024: Start Registration
View Submission - EcoSta 2025
A0866
Title: Local Frechet regression with spherical predictors Authors:  Chang Jun Im - Seoul National University (Korea, South) [presenting]
Jeong Min Jeon - Ewha Womans University (Korea, South)
Abstract: The purpose is to study nonparametric regression methods for non-Euclidean responses and predictors. Specifically, responses taking values in a general metric space and predictors taking values on a general sphere are covered. Both local constant and local linear estimators are proposed, and their asymptotic properties are derived, including consistency and convergence rates. Simulations and real data applications with various response types demonstrate that the methods proposed outperform existing approaches.