 This paper proposes a new method for predicting the chance of survival in nasopharyngeal cancer patients. It combines a highly accurate machine learning model with explainable artificial intelligence to stratify patients into low and high chance of survival groups. The predictive performance of the model was compared with a state-of-the-art algorithm extreme gradient boosting, XGBoost, and it showed similar accuracy. Additionally, two techniques local interpretable model agnostic explanations, LIME, and Shapley Additive Explanations, SHAP, were used to provide personalized protective and risk factors for each patient and uncovered some novel non-linear relationships between input features and survival chance. This method shows promise in providing more accurate predictions of survival in nasopharyngeal cancer patients, which could be useful for effective treatment planning and informed clinical decisions.