Comparison Between Two Common Estimation Methods in Nonparametric Regression Model With Fourier Series Approach (Case Study: Life Expectancy Data)

Estimation plays a crucial role in performing modelling through regression analysis, weather for determining parameters or constructing the model. The OLS and PLS methods are two common estimation methods used for parameter estimation in the nonparametric regression model with Fourier series approach. Many researchers have utilized these methods for parameter estimation in such a model. However, in the previous research, there is no research comparing these methods. Therefore, the purpose of this research is to evaluate and compare the OLS and PLS methods for parameter estimation in the nonparametric regression model with Fourier series approach. For application, we utilize life expectancy data from East Java Province in Indonesia for the year 2022 to assess the performance of these methods. According to theoretical findings, the OLS method depends solely on the number of oscillation parameter H . Meanwhile, the PLS method considers both H and the value of the smoothing parameter λ_j . The optimal values of H and λ_j are determined using the GCV method. In applying the life expectancy data, we obtain that the OLS method is the best estimation method based on the R² and MSE values.