Machine Learning for Survival Prediction in Breast Cancer

In the last few years, machine learning revealed an important instrument to support decision making in oncology. In this manuscript, an application is presented about the use of several machine learning algorithms for the prediction of the survival rate of breast cancer patients. Before presenting the results, the manuscript contains a rather basic introduction to the foundations of machine learning, that can be useful for medical doctors that are not expert in the area. The experiments were carried on using the well-known 70-gene signature dataset for breast cancer. The presented results highlight that genetic programming has interesting advantages compared to other machine learning algorithms, both in terms of prediction accuracy and in terms of model interpretability.