Predicting Breast Cancer Outcome under Different Treatments by Feature Selection Approaches

Abstract

Gene expression data have been used in many researches to help reveal the underlying mechanism of many diseases. In this study, we applied feature selection techniques on breast cancer patients in the METABRIC Study to predict whether patients will be disease free or not, under different treatments. Our models for prediction are of high performance, thus, the genes in those models might help reveal the mechanism of the disease, and these potential biomarkers can become targets for new therapies.