Rachel Lee
Research Mentor: Alexander Rodríguez
Mentor Department: Computer Science and Engineering, Engineering
Author(s): Rachel Lee, Ruipu Li, Alexander Rodríguez
Session: Session 4 (1:00 PM – 1:50 PM)
Presentation Type: Poster 23
Abstract
Infectious disease forecasts play a critical role in public health decision making, especially when agencies need to allocate limited resources under uncertainty. In the United States, several research groups and public health institutions produce disease forecasts, and recent efforts such as CDC forecasting hubs have aimed to coordinate these predictions to support public health planning. This research is also motivated by real-world public health needs, as our main job is to produce and submit weekly forecasts for different diseases to the U.S. Centers for Disease Control and Prevention (CDC), which can inform operational decisions such as adjusting hospital bed capacity or staffing levels in anticipation of increased patient demand. My main role was maintaining the submission pipeline and developing evaluation modules to monitor performance. To support this goal, I developed technical skills in tools such as Linux-based workflows, SSH, and Python data processing while also studying prior research on the retrospective evaluation of epidemic forecasting models. These studies assess model performance after true outcomes are observed and compare forecasting approaches using metrics such as the Weighted Interval Score (WIS) and Coverage, which measure predictive accuracy and uncertainty respectively. Beyond technical development, this research demonstrated how data science and computational skills can directly impact real world decision making. By contributing forecasting outputs that support CDC public health planning, this work highlights the importance of data-driven methods in healthcare and the role of uncertainty quantification in informing policy decisions.
