1. UKM Learning and Research Repository
2. Master Thesis / Tesis Master
3. Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat

Title:	Tabu-based non linear great deluge algorithm for course timetabling problems
Authors:	Rohazlin Md Yassin (P43892)
Supervisor:	Salwani Abdullah, Assoc. Prof. Dr.
Keywords:	Algorithms Universiti Kebangsaan Malaysia -- Dissertations Dissertations, Academic -- Malaysia
Issue Date:	10-Jun-2011
Description:	University Course Timetabling is a complex problem and cannot be dealt with by using only a few general principles. Every academia needs to produce the timetables in every semester or twice yearly. The complicated relationships between timeslots, courses and classrooms make it difficult to obtain a high quality timetable. Many approaches in the literature have addressed this problem. The research work presented in this thesis intends to build upon the state of the art to improve the university course timetabling problem. This work presents a hybridising of tabu search and a Non-Linear Great Deluge algorithm called the Tabu-based Non-Linear Great Deluge algorithm. The Non-Linear Great Deluge algorithm is a modified version of the Great Deluge algorithm. The major different between the two algorithms is the decay rate of the water level where the decay rate of the water level for the Non-Linear Great Deluge algorithm is non-linear, whilst in Great Deluge algorithm, the water level decreases steadily in a linear fashion. Tabu Search are useful to help the search move away from previously visited portions of the search space and thus perform more extensive exploration. Characteristic of Tabu Search could further prevent the local search from wondering in the local optima and indirectly permits for better timetable to be produced. The proposed approach uses a short term memory in tabu search in order to improve the searching process of the Non-Linear Great Deluge algorithm. The tabu list used in this research is in the ranges of two to eight in length. An event that has improves the penalty cost of the timetable is then added in the tabu list for a certain number of iterations in order to avoid cyclic moves. This indirectly allow in generating better course timetabling solutions. We test the proposed approach on the standard benchmark datasets extracted from the Meta-heuristics Network and evaluate the approach performance using standard proximity cost. The experiment results demonstrate that the proposed hybrid approach is capable to produce better results compared to the original Non-Linear Great Deluge algorithm, and competitive with others in the literature.,“Certification of Master’s/Doctoral Thesis” is not available,Master Information Technology
Pages:	97
Call Number:	QA9.58.R636 2011 3 tesis
Publisher:	UKM, Bangi
Appears in Collections:	Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.