Partitioning Students into Cohorts During COVID-19

摘要

The COVID-19 pandemic has forced educational institutions to make significant changes to safeguard the health and safety of their students and teachers. One of the most effective measures to reduce virus transmission is partitioning students into discrete cohorts. In primary and middle schools, it is easy to create these cohorts (also known as "learning groups"), since students in each grade take the same set of required courses. However, in high schools, where there is much diversity in course preferences among individual students, it is extremely challenging to optimally partition students into cohorts to ensure that every section of a course only contains students from a single cohort. In this paper, we define the Student Cohort Partitioning Problem, where our goal is to optimally assign cohorts to students and course sections, to maximize students being enrolled in their desired courses. We solve this problem by modeling it as an integer linear program, and apply our model to generate the Master Timetable for a Canadian all-boys high school, successfully enrolling students in 87% of their desired courses, including 100% of their required courses. We conclude the paper by explaining how our model can benefit all educational institutions that need to create optimal student cohorts when designing their annual timetable.