Meerkat Swarm Optimization Algorithm for Real World University Examination Timetabling Problem

Abstract

Timetabling is a discrete and combinatorial optimization problem. It is handled usually with heuristic and/or artificial intelligent methods. Examination Timetabling Problem (ETP) is a part of the Timetabling problem that comprises allocating a set of exams in a finite period of time while satisfying a set of constraints of various types. Examination timetabling problem has still been studying due to their wide applications. This paper presents Meerkat Swarm Optimization (MSO) technique which is a new intelligent method that has been proposed for the examination timetabling problems, which inspired by the behavior of meerkat swarms in nature. The MSO algorithm simulates cooperative behavior such as caring and foraging of meerkats. Meerkats in the MSO algorithm has its own leaders (called alpha) and they divided into two sub-groups, one for foraging while other stays as a babysitter for pups in burrows. With these strategies, it achieves the diversity in solutions and the individuals in the new algorithm explore and exploit the search space more efficiently. The study deal with the uncapacitated examination timetabling problem. The proposed algorithm is tested on seven benchmark examination timetabling instances. Experimental results prove that the proposed algorithm can produce a promising set of solutions for each examination timetabling instance by comparing our proposed algorithm with Bee Colony Optimization (BCO) and BCO with tournament selection strategy. In addition, an overall comparison examination has been made with the best known results on uncapacitated timetabling benchmark datasets, which well across all the problem instances shows that our approach is competitive and works well in solving this type of problems