Zombie survival optimization in solving university examination timetabling problem

Timetabling is a task of assigning a set of events into a set of resources and satisfying predefined constraints. University timetabling is one of the most stdied timetabling problems among the timetabling domains. It is also a time consuming administrative task that need to be performed in all the...

Full description

Saved in:
Bibliographic Details
Main Authors: Fong, Cheng Weng, Asmuni, Hishammuddin, Leong, Pui Huang, Sam, Yet Huat, Pang, Yee Yong, Sim, Hiew Moi
Format: Conference or Workshop Item
Published: 2022
Subjects:
Online Access:http://eprints.utm.my/id/eprint/98861/
http://dx.doi.org/10.1109/I2CACIS54679.2022.9815494
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Timetabling is a task of assigning a set of events into a set of resources and satisfying predefined constraints. University timetabling is one of the most stdied timetabling problems among the timetabling domains. It is also a time consuming administrative task that need to be performed in all the academic institutions as there are many constraints needed to be considered. In this study, a zombie survival optimization (ZSO) has been applied to address university examination timetabling problem. The underlying idea of ZSO is based on the foraging behavior of zombies, where zombies represent searching agents (solutions) in searching for antidote (optimal solution). There are three modes in ZSO, namely exploration mode, hunter mode and human exploitation mode where zombies explore for solutions (randomly) in exploration mode, explore towards a human (promising search region) in hunter mode and turn into human to search (exploitation) for local optimum. The ZSO is tested on Carter's university un-capacitated examination benchmark dataset and results demonstrated that ZSO is capable of producing promising quality of solutions when compared with the published methods in the literature. In fact, ZSO managed to record new best-known results on 3 instances of the dataset.