The Curve Estimation of Combined Truncated Spline and Fourier Series Estimators for Multiresponse Nonparametric Regression

Nonparametric regression becomes a potential solution if the parametric assumption is too restrictive while curve assumed to be known. In multivariable nonparametric regression, pattern of each predictor variable’s relationship with response variable not always same; thus, combined estimator recommended. addition, modeling sometimes involves more than one response, i.e., multiresponse situations. Therefore, we propose new estimation method performing estimator. The objective estimate using truncated spline and Fourier series estimators for regression. proposed model obtained via two-stage estimation: (1) penalized weighted least square (2) square. Simulation data sample size variation different error variance were applied, where best satisfied result through large small variance. Additionally, application real dataset human development index indicators in East Java Province, Indonesia, showed that had better performance uncombined estimators. Moreover, an adequate coefficient determination indicated successfully explained variation.