Uncovering Socio-Economic and Demographic Determinants of Global Suicide Rates

In this data mining project, I conducted a comprehensive cross-country analysis to identify key socio-economic and demographic factors influencing global suicide rates. Leveraging a panel dataset from Kaggle containing over 27,000 records across 1985–2016, I applied statistical modeling and machine learning techniques—including linear and non-linear regression, regression trees, boosting, and support vector regression—to uncover complex relationships between suicide rates and variables such as GDP, HDI, age, gender, and population. The analysis revealed that suicide rates were notably higher among males and older adults, and inversely associated with GDP and population. Interestingly, HDI showed a positive correlation with suicide rates, prompting a deeper discussion into socio-economic metrics and mental health outcomes. I utilized R for the entire workflow, incorporating packages such as caret, glmnet, rpart, gbm, and e1071 to perform data cleaning, transformation, visualization, model building, and performance evaluation (MSE, MAE, AIC, and adjusted R²). Forward variable selection, interaction modeling, and normalization techniques enhanced the robustness of the insights. The project’s findings offer valuable guidance for public health policy and resource allocation, particularly in developing targeted mental health interventions for high-risk demographics and economically challenged regions.