In a recent study, five different machine learning algorithms were leveraged to create a single distinct algorithm known as a stacked algorithm for the prognostication of overall survival outcomes in nasopharyngeal cancer patients. The performance of this algorithm was compared to another state-of-the-art algorithm called extreme gradient boosting (XGBoost). To make the predictions explainable and interpretable, the Local Interpretable Model Agnostic Explanations (LIME) and Shapley Additive exPlanations (SHAP) techniques were used. The resulting model may aid in personalized chance of survival stratification for patients, allowing for tailored treatment intensity. Data from the National Cancer Institute database through the Surveillance, Epidemiology, and End Results (SEER) Program of the National Institutes of Health were used.
The study’s lead author is Alabi Alabir, a Ph.D. student in the Department of Oncology-Pathology, Karolinska Institutet, Stockholm, Sweden.
The research was conducted independently and did not involve any outside companies.
