Steve Broll

---

I am a Postdoctoral Fellow in the Michigan Institute for Data & AI In Society (MIDAS). I am also a member of the Irina Stats Lab run by Dr. Irina Gaynanova , and the Michigan Multi-Pollutant Epidemiology Group (MPEG) run by Dr. Sung Kyun Park . I am a statistician specialized in the development of methodology for biomedical data involving the modeling of disease onset or progression over time. I am particularly interested in data that is high-dimensional and time-varying in both outcome and predictors. My current research focuses on utilizing time-varying omic data (metabolomic, microbiomic, exposomic) together with wearable continuous glucose monitoring data to model onset and progression of type-2 diabetes. Previously, I was a Postdoctoral Associate in the Division of Nutrional Sciences (DNS) and Center for Precision Nutrition and Health (CPNH) at Cornell University, advised by Dr. Saurabh Mehta and Dr. Julia Finkelstein . I obtained my PhD in Statistics in 2025 from Cornell University, where I was very fortunate to be advised by Dr. Martin T. Wells, Dr. Sumanta Basu, and Dr. Myung Hee Lee. At Cornell I was a member of the Statistical Modeling of Complex Systems (SMoCS) Lab, and I was a T32 Fellow in AI and Precision Nutrition (AIPrN). My dissertation work centered on variable selection for time-varying outcome-guided longitudinal omics data, developing a model framework called PROLONG that first-differences and vectorizes data before fitting a group lasso and graph-Laplacian double-penalized linear model. I obtained my B.S. in Statistics in 2020 from Texas A&M University where, supervised by Dr. Irina Gaynanova, I developed popular R package `iglu` and a corresponding Shiny Application that produces metrics and visualizations for continuous glucose monitoring (CGM) data.