Predicting factors in financial loss among Malaysian scam victims using machine learning

This study presents an innovative educational approach to scam prevention by using logistic regression and decision tree models to identify key predictors of financial loss among 394 Malaysian scam victims. Emotional harm, age, and cybersecurity knowledge emerged as the most significant factors, with emotional harm being the strongest predictor of these factors. The decision tree model demonstrated superior accuracy and interpretability compared to logistic regression, making it a practical tool for educational use. By integrating data science with digital literacy, this research supports the development of targeted learning modules and public awareness strategies. The findings emphasize the use of machine learning to enhance risk education, empower self-assessment, and inform evidencebased interventions aimed at reducing scam victimization in Malaysia.