The global rise in obesity, particularly among adults aged 25 to 45, presents significant public health challenges. Despite numerous studies on the behavioral and biological factors of obesity, few have leveraged machine learning to predict weight management behaviors in this critical age group. This study applies and compares multiple machine learning models—including decision trees, random forests, logistic regression, and support vector machines—to identify key factors influencing weight management and to develop predictive models for intervention. Data were collected from 843 participants through comprehensive questionnaires on dietary habits, physical activity, and family health history. Among the models tested, the random forest achieved the highest accuracy (98.7%), but the decision tree was selected for its superior interpretability (97.3% accuracy), making it more suitable for clinical decision-making. Age, body weight, and family history of obesity emerged as the most influential predictors. These findings support the integration of interpretable machine learning models into clinical practice and public health initiatives to identify high-risk individuals and tailor preventative strategies. Future research should incorporate psychological and environmental factors to further enhance predictive accuracy.