<![CDATA[The Human Development Index (HDI) is an important indicator for measuring the quality of development in a region. This study compares two nonparametric regression approaches, namely truncated spline regression and Gaussian kernel regression, in analyzing the factors influencing HDI in Indonesia in 2024. The independent variables used include Expected Years of Schooling (HLS), Mean Years of Schooling (RRLS), and the percentage of the poor population (PPM). Nonparametric regression is chosen for its ability to capture complex relationships between variables without strict linearity assumptions. The results show that both methods effectively model the relationship between the variables and HDI. Truncated spline regression performs better in detecting structural changes, while kernel regression is more flexible in capturing smooth relationships. Model evaluation using the coefficient of determination (R²) and mean squared error (MSE) indicates that truncated spline yields an R² of 92.79% and an MSE of 1.8617, while Gaussian kernel regression results in an R² of 82.25% and an MSE of 3.6837. Therefore, truncated spline regression proves to be more accurate in modeling the relationship between determining factors and HDI, and it can serve as a more suitable alternative for analyzing complex and nonlinear patterns in human development policy research.]]>
Copyrights © 2025
