The purpose of this study was to model the Gross Regional Domestic Product (GRDP) in Indonesia using nonparametric spline regression and Fourier series. The data used includes the GRDP of 34 provinces in Indonesia, with predictor variables such as labor force participation rate, foreign direct investment, local revenue, minimum provincial wage, and human development index. In the nonparametric spline regression, the determination of the optimal knot points is performed with one to three knots, and the optimal knot points are obtained by minimizing the Generalized Cross Validation (GCV) value, resulting in three optimal knot points. In the nonparametric Fourier series regression, calculations are performed for one to three oscillation points, and the optimal oscillation points are obtained based on the minimum GCV value, resulting in three optimal oscillation points. Therefore, the best nonparametric spline regression model utilizes three knot points, while the Fourier series model utilizes three oscillation points for modeling the GRDP in Indonesia. Keywords: Fourier Series; nonparametric regression; oscillation point; spline
Copyrights © 2023