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

Title:	An enhancement of the Cuckoo search (CS) algorithm using the Centroidal Voronoi Tessellations (CVT)
Authors:	Shatnawi Moaath Mohammad Amin (P53706)
Supervisor:	Mohammad Faidzul, Dr.
Keywords:	Dissertations, Academic -- Malaysia Cuckoo search Centriodal Voronoi Tessellations
Issue Date:	31-Jul-2011
Description:	A recent meta-heuristic algorithm, the Cuckoo Search (CS), was developed by Yang and Deb to solve optimization problems. It was inspired by the obligate brood parasitism of some cuckoo species by laying their eggs in the nests of other host birds. The performance of the CS algorithm can further be improved by strategically selecting the starting positions of the nests instead of the standard random selection. This work suggests the use of positions generated from the Centroidal Voronoi Tessellations (CVT) as the starting points for the nests of the original CS algorithm to reduce the number of evaluations. CVT is a Voronoi Tessellation of a set such that the generators of the Voronoi sets are simultaneously their centers of mass. This version of enhanced CS algorithm is called CVTCS. The performance of CS algorithm initialized using this method is compared with those generated from the standard CS algorithm (without the CVT) on welded beam design problem as a real optimization problem and several standard benchmark testing function which are Ackley’s, De Jong’s, Easom’s, Shubert’s, Rastrigin’s, Schwefel’s, Rosenbrock’s and, Mechelewicz’s. The results suggest that the initialization of CS algorithm using the CVT improves its performance especially for functions with high-dimensional input spaces.,Master
Pages:	92
Call Number:	QA166.8 .S535 2011 3
Publisher:	UKM, Bangi
URI:	https://ptsldigital.ukm.my/jspui/handle/123456789/476318
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.