Chronic diseases such as type II diabetes have become increasingly prevalent over the past few decades. If left untreated, these
conditions lead to many complications that tremendously affect patients’ quality of life, and also create a large economic burden to the healthcare industry and governments. The goal of this project is to model the disease trajectory, and identify risk factors and interventions that can change a patient’s disease state. We use techniques in machine learning including latent variable modeling and time series analysis, as well as extensive medical feature design, for this task.

As the first step in the project, we focus on improving accuracy of early prediction of type II diabetes, by automatically discovering risk factors from administrative healthcare records. Both linear and nonlinear models are explored, and currently we focus on time series modeling of patient state.


Moore Sloan Poster Research Project