Advanced analytics to predict survivability of breast cancer patients

Cancer is major burden of disease worldwide. Amongst women, breast cancer is the most common cancer and primary cause of death followed by heart diseases. With increasing breast cancer cases and technological improvements, cancer care institutions and registries have collected large volumes of data in various formats. Unfortunately, these repositories are not easily accessible and the stored formats are difficult to analyze. We propose an end-to-end process through which such data can be cleansed, integrated and presented in the form of interactive dashboards. This provides a comprehensible view of over forty years of data consisting of over one million records with provisions to slice this data along several dimensions. Additionally, developing a breast cancer predictive model that predicts survival months for diagnosed patients is proposed. The outcomes of different modeling techniques along with assessing the impacts of retraining the predictive model is observed in the experimentations conducted for this research.