Better Class Schedules for Better University Education - Mathematical Modelling Helps

What is it about?

Class-scheduling, also known as timetabling, in universities is a difficult combinatorial problem involving allocation of various academic resources including faculty staff, class rooms, time periods. Well constructed class schedules enable proper utilization of these resources and effective delivery of instructions in ways that are more convenient to both students and faculty. Mathematical programming can help in achieving these goals. In this study we demonstrate this through an actual application of a sequential integer programming approach to develop class-schedules at the college of Business and Economics, Bahir Dar University, Ethiopia.

Featured Image

Why is it important?

The model along with an excel based database are the core components of the system which we used to construct official schedules for the college’s first and second semesters of its 2013–2014 academic year. These initial implementations have demonstrated that using our model (Vs manual construction of schedules) can achieve significant improvements in several dimensions of schedule quality. We generated complete and conflict-free schedules and reduced the time required to construct schedules from one week to less than an hour. We also improved the percentage of morning classes from 50 to 60 percent and significantly reduced room switching. The teaching loads generated achieve a more balanced distribution for instructors, and lectures are spread more evenly across the weekdays per section. Besides these practical values, our sequential integer programming approach is a unique approach to the university class scheduling problem to make manageable, through decomposition into smaller sub-problems, an otherwise large, complex problem .