Session

session6

Location

Space Dynamics Laboratory, Auditorium Rm C

Start Date

5-9-2022 11:10 AM

End Date

5-9-2022 11:20 AM

Description

The COVID-19 pandemic caused a lot of disruption in the academic world. Many schools and universities shut down entirely and transitioned to online learning. As BYU made the transition back to in-person learning, the administration needed to know if there was in-class transmission happening on campus in order to regulate restrictions and keep students and faculty safe. Using demographic information about the students, we built an XGBoost model that produces an estimated probability of testing positive for each student. We incorporated engineered variables from the demographic information as well. We evaluated model fit using metrics such as AUC. Using a simulation-based approach, we developed a method that identifies possible spots of in-class transmission on campus with these probabilities. We simulated semesters under the null hypothesis, that there is no in-class transmission, and compared them to the observed positivity rates to find a p-value for each group (section, course or major). Given that our p-values are dependent and non-uniform, we developed a simulation-based method to control the false discovery rate for dependent, non-uniform p-values. We performed simulation studies to test the accuracy of our method in terms of FDR and sensitivity. This method successfully identifies problematic groups with moderate levels of transmission.

Available for download on Tuesday, May 09, 2023

Share

COinS
 
May 9th, 11:10 AM May 9th, 11:20 AM

Progress Report: COVID-19 Hotspot Detection in University Campus Settings

Space Dynamics Laboratory, Auditorium Rm C

The COVID-19 pandemic caused a lot of disruption in the academic world. Many schools and universities shut down entirely and transitioned to online learning. As BYU made the transition back to in-person learning, the administration needed to know if there was in-class transmission happening on campus in order to regulate restrictions and keep students and faculty safe. Using demographic information about the students, we built an XGBoost model that produces an estimated probability of testing positive for each student. We incorporated engineered variables from the demographic information as well. We evaluated model fit using metrics such as AUC. Using a simulation-based approach, we developed a method that identifies possible spots of in-class transmission on campus with these probabilities. We simulated semesters under the null hypothesis, that there is no in-class transmission, and compared them to the observed positivity rates to find a p-value for each group (section, course or major). Given that our p-values are dependent and non-uniform, we developed a simulation-based method to control the false discovery rate for dependent, non-uniform p-values. We performed simulation studies to test the accuracy of our method in terms of FDR and sensitivity. This method successfully identifies problematic groups with moderate levels of transmission.