MODELING ENVIRONMENTAL QUALITY INDEX IN INDONESIA WITH MULTIVARIATE ADAPTIVE REGRESSION SPLINES

Aims: This study aims to analyze the factors influencing Indonesia’s Environmental Quality Index (IKLH) using nonparametric regression methods, specifically Multivariate Adaptive Regression Spline (MARS) and multipredictor truncated spline. Study Design: A quantitative study applying nonparametric regression analysis. Place and Duration of Study: Data were collected from the Central Bureau of Statistics and the Ministry of Environment for 34 provinces in Indonesia in the year 2022. Methodology: The study used secondary data comprising predictor variables such as Human Development Index (HDI), population density, proper sanitation, poverty percentage, and Gross Regional Domestic Product (GRDP). Two nonparametric regression methods were applied: MARS and multipredictor truncated spline. Model performance was assessed using Mean Squared Error (MSE), Generalized Cross-Validation (GCV), and R² values. Results: The truncated spline method demonstrated superior performance with MSE = 5.63308, GCV = 10.42, and R² = 82.63%, compared to MARS with MSE = 7.685, GCV = 16.014, and R² = 79.3%. The analysis confirmed that HDI, population density, proper sanitation, poverty percentage, and GRDP significantly influence IKLH. Conclusion: The findings highlight the key determinants of environmental quality in Indonesia and emphasize the effectiveness of nonparametric regression approaches in environmental studies. These insights provide a valuable reference for policymakers in designing strategies to improve environmental sustainability.