Suicide is regarded as a public health emergency, which is believed to be mediated by important socio-economic factors that differ between geographical areas. In the current study, the association of 22 socio-economic indicators with total suicide rates was examined, using data obtained from the World Data Bank across six Asian countries. Advanced machine learning techniques such as XGBoost, Random Forest, AdaBoost, Gradient Boosting, and LightGBM were applied. Among these, XGBoost was found to perform best after extensive hyperparameter optimization, with an adjusted R² of 0.718 and RMSE of 1.536 on the training data, and an adjusted R² of 0.773 and RMSE of 3.857 on the test data. For interpretation, Partial Dependence Plots (PDPs) of the XGBoost model were plotted, providing the marginal effects of certain individual factors on the response of interest – suicide rates. This work is intended to contribute to the existing body of knowledge on the socio-economic determinants of suicide that can be used to inform evidence-based interventions.