Universal Local Linear Kernel Estimators in Nonparametric Regression

New local linear estimators are proposed for a wide class of nonparametric regression models. The uniformly consistent regardless satisfying traditional conditions dependence design elements. the solutions specially weighted least-squares method. can be fixed or random and does not need to meet classical regularity independence conditions. As an application, several constructed mean dense functional data. theoretical results study illustrated by simulations. An example processing real medical data from epidemiological cross-sectional ESSE-RF is included. We compare new with best known such studies.