A hybrid clonal selection algorithm with conflict based ststistics for university course timetabling

Abstract

The University course timetabling problem involves the allocation of courses to rooms and timeslots subject to satisfaction of hard and soft constraints. The hard constraints must be satisfied, while the soft constraints are desired to be satisfied. The problem also has an objectice function that need to be maximised. Several methodologies have been used for solving timetabling problem such as the sequential methods, graph coloring, cluster methods, constraint based and meta heuristic methods. The Hybrid Clonal Selection Algorithm with Conflict Based Statistics (Hybrid CLONALG-CBS) was chosen based on CLONALGs’ positive track record in optimization tasks and the ability of CBS in avoiding conflicting value assignments to a variable. The Hybrid CLONALG-CBS start with an initial solution, the initialized solution then undergo selection, cloning and mutation; the mutated solutions are used for the generation of improved solutions. The dataset is from Faculty of Computer Science and Information System, Universiti Teknologi Malaysia. The experimental results showed the Hybrid CLONALG-CBS fared better than the manual method and CLONALG algorithm in timeslot utilization, room utilization, Lecture spread and objective function.