A two-stage stochastic programming approach for a multi-objective course timetabling problem with courses cancelation risk

Abstract We study a university course timetabling problem, where the registration is implemented in two steps: pre-registration and drop/add phases. Since the ultimate timetable is finalized based on the students’ decisions made on the drop/add phase, there may be some courses need to be canceled because of not reaching to the threshold value in terms of total registered students. As a result of cancelations, some undesirable changes may be imposed on the final timetable for the students and the professors as well. In this paper, we ideally arrange course timetable, while considering the courses cancelations risk and the possible changes which may happen after the drop/add period; ultimately the undesirability is minimized in the final timetable. In order to achieve that, we optimize an objective function which consists of seven different objectives combined as a weighted sum function. As regards the solution method, we develop a two-stage stochastic programming model, a heuristic approach and a two-stage separated model where the latter one model the traditional method. The performance of all developed algorithms is then analyzed using randomly generated test instances.