Kernel-Spline Estimation of Additive Nonparametric Regression Model

9th Annual Basic Science International Conference 2019

自然科学（总论）

Hidayat, Rahmat^1^2 ; Budiantara, I Nyoman^2 ; Otok, Bambang Widjanarko^2 ; Ratnasari, Vita^2

Department of Mathematics, Cokroaminoto Palopo University, Palopo, Indonesia^1

Department of Statistics, Faculty of Mathematics, Computation and Data Science, Institut Teknologi Sepuluh Nopember, Surabaya, East Java, Indonesia^2

关键词: Coefficient of determination;    Gaussian kernels;    Generalized cross validation;    Kernel function;    Non-parametric regression;    Performance Model;    Spline kernels;    Unemployment rates;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/546/5/052028/pdf
DOI  :  10.1088/1757-899X/546/5/052028

学科分类：自然科学（综合）

来源: IOP

【 摘 要 】

In this paper, we model the open unemployment rate with the Kernel-Spline model. We investigate and compare the performance of model Kernel-Spline by varying the Kernel function. The performance model has been compared with five Kernel function i.e. Kernel functions Uniform, Epanechnikov, Quartic, Gaussian, and Triweight. For these models, we conducted a comparison based on actual data sets, the unemployment rate in East Java. The best model was chosen based on the Generalized Cross Validation value and the coefficient of determination criteria. The empirical results obtained have shown that Spline-Kernel model by using the Gaussian Kernel better than other models.