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

Title:	Patch-levy-based bees algorithm for static and dynamic function optimization
Authors:	Hussein Wasim Abdulqawi Ahmed (P56228 )
Supervisor:	Shahnorbanun Sahran, Associate Professor Dr.
Keywords:	Mathematical optimization. Patch-Levy model
Issue Date:	26-Oct-2015
Description:	The Bees Algorithm (BA) is a swarm-based metaheuristic algorithm inspired by the natural behavior of honeybees when foraging for food. Since its invention, more than one version of the algorithm has been proposed. However, the convergence speed of BA to the optimal solution still needs further improvement and it also needs a mechanism to obviate getting trapped in local optima. Initial locations of foragers relative to the optimal resource (target) can significantly affect the quality of the resource reached and the speed of convergence to the optimal target. Therefore, a novel initialization algorithm based on the patch concept and Levy flight distribution, called the Patch-Levy-based Initialization algorithm (PLIA) has been introduced and incorporated into Basic BA to propose an enhanced BA variant that is denoted by PLIABA. However, for most real-world problems, the local and global capabilities are also required to be improved to enhance the exploitativeness and exploration capabilities of BA, respectively, thus supporting quick convergence to the optimal solution. Consequently, a new algorithm for the local search part called the Greedy Levy-based Local Search Algorithm (GLLSA), which models Levy looping search for the exploitation of patches has been adopted. The global search has also been enhanced and performed based on the patch-Levy model adopted in the initialization procedure. Based on the improved initial, local and global stages, a new variant of BA called the PatchLevy-based Bees Algorithm (PLBA) has been proposed. Many real-world problems are of dynamic nature where the optimal solution changes over time. Thus, it is very crucial to adopt optimization algorithms that are able to track the optimum in such problems. The PLIA-BA is evaluated on several widely used high-dimensional standard static benchmark functions and compared with some other variants of the BA and state-ofthe-art bee swarm-based algorithms. The results indicate that PLIA-BA significantly outperforms other BA variants and state-of-the-art variants of the Artificial Bee Colony (ABC) algorithm in terms of solution quality, convergence speed, and success rate. Then, PLBA is evaluated on several more challenging static benchmarks that include non-separable and composition functions. Additionally, the performance of PLBA is evaluated on multilevel image thresholding based on Kapur’s entropy and Otsu’s criteria using a set of widely used real images. The results are compared with those from the PLIA-BA, other BA variants and other state-of-the-art algorithms. The results show that PLBA significantly outperforms other BA variants and some other algorithms. To evaluate the performance of PLBA and other BA versions in the dynamic environment, a set of recently dynamic benchmark functions are used. The experimental results show that PLBA outperforms other BA variants and some of other state-of-the-art algorithms, and achieves,Ph.D
Pages:	248
Call Number:	QA402.5 .H847 2015 3
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.