FURI | Spring 2021
Model-on-Demand (MoD) Estimation for Behavioral Intervention Optimization
This project explores the use of Model-on-Demand (MoD) estimation to model noisy, nonlinear time series data. MoD estimation has been examined for two applications: (1) a continuously-stirred tank reactor and (2) walking behavior from HeartSteps, a microrandomized study that aims to encourage physical activity. While the former demonstrates that MoD can be applied to noisy, nonlinear data, the latter explores MoD’s effectiveness in more complex systems, accounting for environmental factors such as weather and time of day. Better models created through MoD will allow researchers to provide more accurate behavioral health interventions.
Hometown: Chandler, Arizona, United States
Graduation date: Spring 2021