A Great Deluge Algorithm for University Course Timetabling

Abstract

Here we propose a two-stage approach to address the University Course Timetabling Problem. In the first stage, we utilize an existing algorithm called Tabu Search with Sampling and Perturbation to generate a feasible solution, and in stage two, we use a Great Deluge algorithm to improve the quality of this feasible solution. The proposed methodology was tested on a recent instance of real university timetabling data from Universiti Malaysia Sabah. Experiments were conducted to determine the most suitable GD algorithm parameter values that resulted in optimal performance. Additionally, the performance of the GD algorithm was compared with that of a Genetic Algorithm and was found to achieve a lower minimum and lower average cost values.