An heuristic genetic algorithm for strategic university tuition planning and workload balancing

Abstract

The typical tuition planning challenge that universities face is to optimize students’ and teachers’ timetables on a weekly basis. While crucial for operational arrangement of tuition, such short-term planning is often done in relative isolation from longer-term strategic considerations, such as organizational competence development. This article presents an heuristic method, based on genetic algorithms, for long-term strategic planning of tuition for multi-year university degree programmes, at the semester level. The proposed method attempts to produce such a tuition schedule that satisfies given pre-requisite knowledge constraints, while honouring teacher-specific preferences, such that the workload across the teachers is as uniform as possible. The proposed method accommodates many real-world constraints, and can be applied to several distinct student groups advancing in parallel. As an application, the proposed method is used for optimizing the tuition schedule of a mechanical engineering degree programme at a Finnish university.