Improving Finite Samples Performances in Nonparametric Functional Regression by Using Weighted Pseudo-Metrics
摘要
The choice of the pseudo-metric plays a crucial role in functional nonparametric regression as it has a direct impact on the type of models one considers. In this work, the idea of weighted pseudo-metric is introduced and discussed. The effects and the performances of the choices of pseudo-metric used are evalulated by means of a Monte Carlo study.