Preference-based assignment of university students to multiple teaching groups

A successful approach to the student-scheduling problem is presented here. This problem arises naturally when courses and classes must be offered in such a large number of multiple teaching groups that some of these are timetabled in parallel, i.e. simultaneously. This is the case, for example, at the School of Business, Economics and Social Sciences of Universität Hamburg, Germany. Here, 3,735 students have to be assigned in a real-life scenario to 48 courses in 300 multiple teaching groups such that no student has a time clash and students’ enrolments as well as group capacities are met. Students’ satisfaction with regard to individually specified preferences for various aspects of the scheduling can be used as the objective which yields a multi-criteria decision problem. The resulting mixed-integer programme is modelled in GAMS and typical instances can be solved within minutes using the commercial CPLEX solver. Furthermore, the integration of the optimisation approach into the existing registration process using a campus-management software is discussed.