Alternative Functionals Estimation Based on Index Models and Kernel Approach

Abstract

This work explores the utilization of double cross-validation methods for determining the optimal bandwidth in kernel regression using single index-models. Kernel regression is a non-parametric technique widely employed in various fields, particularly in smoothing noisy data. The bandwidth parameter plays a crucial role in kernel regression, controlling the smoothness of the estimated function. Selecting an appropriate bandwidth is essential for achieving accurate and robust model performance. Traditional approaches to band- width selection often rely on heuristic methods or fixed rules, which may not be optimal for all datasets (...)