Data Mining: Medical Data

Member: Cameron Loader

Advisor: Dr. Han

Project Homepage:
Click here to view the project homepage

In this project, we are investigating relationship between Hypoglycemia and cardiovascular disease (CVD) by mining patient data in a diabetes database. The latter is obtained from the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK). The database features 1,200 de-identified patient records. The existing work on the relationship between Hypoglycemia and CVD is both controversial and inconclusive. Our project aims to shed more light on the controversial results by investigating Hypoglycemia and CVD relationship using a new approach — data mining using logistic regression. Since logistic regression can be used for both continuous and discrete data, we expect our approach to yield better accuracy in predicting which diabetes features are at higher risk for CVD and finding associations between the features and CVD.