Open Access Senior Thesis
Bachelor of Science
Lisette de Pillis
2020 Cassidy My Huong Le
Diabetes continues to affect many lives every year, putting those affected by it at higher risk of serious health issues. Despite many efforts, there currently is no cure for diabetes. Nevertheless, researchers continue to study diabetes in hopes of understanding the disease and how it affects people, creating mathematical models to simulate the onset and progression of diabetes. Recent research by David J. Albers, Matthew E. Levine, Andrew Stuart, Lena Mamykina, Bruce Gluckman, and George Hripcsak1 has suggested that these models can be furthered through the use of Data Assimilation, a regression method that synchronizes a model with a particular set of data by estimating the system's states and parameters. In my thesis, I explore how Data Assimilation, specifically different types of Kalman filters, can be applied to various models, including a diabetes model.
1Albers, David J, Matthew E Levine, Andrew Stuart, Lena Mamykina, Bruce Gluckman, and George Hripcsak. 2018. Mechanistic machine learning: how data assimilation leverages physiologic knowledge using bayesian inference to forecast the future, infer the present, and phenotype. JAMIA 25(10):1392–1401. doi:10.1093/jamia/ocy106. https: //doi.org/10.1371/journal.pone.0048058.
Le, Cassidy, "Use of Kalman Filtering in State and Parameter Estimation of Diabetes Models" (2020). HMC Senior Theses. 232.