Application of integer programming techniques to facilitate the production of examination timetables at FIAL faculty
Files
Pypaert_76561100_2018.pdf
Open access - Adobe PDF
- 5.65 MB
Details
- Supervisors
- Faculty
- Degree label
- Abstract
- Besides being an omen of hard times to come for the students, examination timetables are the result of a long and tedious work by the faculty’s staff who is responsible for their generation. From the enrolments until the delivery of the first draft to the students, up to one whole month of work can be needed by the employees of the department. This last statement is far from being a criticism of their aptitude given the complexity to solve such problems even when using considerable computational power. In fact, the essence of the problem lies in the attribution of exclusive resources contained in multiple sets (students, professors, time slots, rooms) to events that are meant to take place (the exams) while taking care of various constraints and requirements that have to be fulfilled to implement the solution in real terms. These constraints vary from institution to institution even if all agree on the impossibility to take two exams simultaneously. When this thesis was written, most of the operations regarding the creation of examination timetables were performed manually at FIAL faculty. Our goal is to determine if the use of integer programming to transpose the problem is doable, facilitates the generation of solutions and improves their quality in comparison with the manual approach. Based on the observations of the experiment, some future points of research will be suggested.