Prediction Model of Breast Cancer Survival Months: A Machine Learning Approach

Breast cancer is one of the most common cancers in women. Because of the importance of early diagnosis in treatment success, the number of routine check-ups has recently risen. The survival rate is one of the top concerns of patients following a diagnosis. Current methods are to find survival rates using traditional statistical approaches by examining people of similar conditions. This study is one of a handful that looked at forecasting breast cancer survival time using Machine Learning (ML) techniques. Using data from 4024 patients from the NIH SEER program and a Random Forest classifier, we were able to estimate breast cancer survival time within a two-year window with up to 72% accuracy. It was also discovered that the patient's age, race, and marital status are all strongly associated with anticipated survival time. This study is a step toward broader use of ML science in clinic-generated data processing, ultimately improving current practices and opening new avenues for challenging medical research problems.