This study assessed the application of machine learning to predict quality of life (QoL) and multidimensional health outcomes in cancer survivors. Researchers used longitudinal data from 256 survivors in South Korea which included sociodemographic and clinical measures, as well as standardized measures of self-management strategies. These data were analyzed using several ensemble machine learning models and validated with repeated stratified K-fold procedures. The XGBoost model demonstrated the best performance with high accuracy for global QoL and particularly strong within the mental and social domains, while having lower predictability with physical health. Key determinants revealed using explainable AI methods included activity-based coping, proactive problem-solving and self-implementation strategies, while age, religion, and income were identified as domain-specific factors. The authors concluded that explainable AI provided not only clarity on which features had the strongest influences for outcomes, but also indicated baseline coping strategies were more influential in predicting long-term QoL than any short-term change.