<![CDATA[Under-five mortality remains a critical indicator of public health performance, reflecting a country’s socio-environmental progress and overall quality of life. However, studies examining the relationship between basic drinking water, sanitation, and under-five mortality in Southeast Asia remain limited, particularly those that integrate panel data with regularized machine learning to address multicollinearity among development indicators. The objective of this study is to analyze the association between basic drinking water access, basic sanitation access, and under-five mortality across 11 Southeast Asian countries from 2003 to 2023, while controlling for GDP per capita, health expenditure, urban population, and DPT immunization. The method used in this study is a balanced panel data approach combined with regularized regression models, including Ridge Regression, LASSO, and Elastic Net. A conventional panel model is selected through panel specification tests, while regularized models are evaluated using nested cross-validation, with Leave-One-Country-Out Cross-Validation in the outer loop and internal 10-fold cross-validation for parameter tuning. Bootstrap inference is applied to the selected regularized model. The novelty of this study lies in the integration of panel data, regularized machine learning, cross-country validation, and bootstrap inference in under-five mortality modeling. The results show that Ridge Regression achieves the best performance, with an RMSE of 0.4037, MAE of 0.3360, and R² of 0.8193. Basic sanitation access, GDP per capita, urban population, and DPT immunization are negatively associated with under-five mortality. These findings imply that reducing under-five mortality requires integrated policies that improve sanitation, expand immunization, strengthen the economy, and ensure equitable access to basic services.]]>
