Data-Driven Modeling to Improve Understanding of Human Behavior, Mobility, and Disease Spread

Models of disease dynamics are important tools used to predict the numbers of cases and deaths over time and to support policymakers as they prepare for and respond to infectious disease outbreaks. However, despite significant advances, many models still lack realistic representations of human behavior and mobility, which are key drivers of disease spread. Without accounting for the complexity of human behavior, models are limited in their ability to make accurate predictions, especially over longer time horizons. This project investigates the inclusion of realistic human behavior and mobility in models of disease spread to 1) better explain the different ways that humans respond to disease outbreaks, 2) improve predictions of infectious disease spread, and 3) help to prescribe the most effective mitigation policies. The investigators use publicly available data so that models can be rapidly deployed for any county or state in the U.S. to predict and mitigate future outbreaks of infectious respiratory diseases (e.g., COVID-19, seasonal influenza, measles, and smallpox). These models may provide more timely and accurate predictions to help the general public, key institutions, and policymakers anticipate what is to come and provide support for evidence-based policy making. This project will support professional development opportunities for early-career researchers and training opportunities for a postdoctoral researcher, graduate, undergraduate, and high school students in the Aspiring Scientists Summer Internship program.

The researchers will use a data-driven approach to explain the spatio-temporal variations in the behavioral response to a disease outbreak. They hypothesize that regional variables such as average income, age, political leaning are associated with spatial patterns of behavioral response, and will leverage very large data sets to mine association rules between such variables and observed behavioral response, including social distancing, stay-at-home behavior, mask usage, and vaccine acceptance. These association rules will be used to develop a novel modeling framework that captures spatio-temporal variations of human response to disease. The proposed modeling framework will be implemented to simulate the spread of COVID-19 using Fairfax County, VA, as a case study. This framework also will be leveraged for prescriptive analytics to find the best course of action in the event of future infectious disease outbreaks. The researchers will simulate and optimize policy measures aimed at mitigating disease spread and minimizing socio-economic impact. This optimization will combine automatic optimization tools with the expertise of researchers in policy, epidemiology, health geography, and psychology.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.