dc.contributor.authorNguyen, Quang Huy
dc.identifier.citationHuy, N. Q. (2010). Adaptive and formal memetic frameworks. Doctoral thesis, Nanyang Technological University, Singapore.
dc.description.abstractOver the recent years, there has been increasing research activities made on improving the efficacy of Memetic Algorithm (MA) for solving complex optimization problems. Particularly, these efforts have revealed the success of MA on a wide range of real world problems. MAs not only converge to high quality solutions, but also search more efficiently than their conventional counterparts. Despite the success and surge in interests on MAs, only restricted theoretical knowledge is available in the field of MA and limited progress has been made on formal memetic frameworks. The key design issue of MA lies in the successful promotion of competition and cooperation between the forces of evolution and individual learning through appropriate configuration of the algorithmic parameters. This can be achieved by 1) applying different parameter configurations to find the most suitable setting or 2) adapting the algorithmic parameters of MA while the search progresses. While the former is straightforward, it could only produce reasonable configurations that are suitable for solving a given problem at hand. Further, such an option may not be feasible due to the high computational cost involved. The latter, on the other hand, may be applied to a wide range of problems more efficiently, but would require suitable online adaptation schemes and frameworks. This thesis provides some insights into the success of MAs and summarizes investigation on the study of effective memetic frameworks. It addresses several important design issues of MAs in the context of continuous optimization problems and proposes two adaptive approaches to efficiently and effectively configure the MA design parameters during the search process. Subsequently, a formal probabilistic memetic framework is presented to highlight the inter-relationship among the MA design issues. Finally, the thesis presents a complexity analysis on the numerical methods as individual learning in MAs.en_US
dc.format.extent193 p.en_US
dc.subjectDRNTU::Engineering::Computer science and engineering::Computer applications::Computers in other systemsen_US
dc.subjectDRNTU::Engineering::Computer science and engineering::Mathematics of computing::Probability and statisticsen_US
dc.titleAdaptive and formal memetic frameworksen_US
dc.contributor.researchCentre for Computational Intelligenceen_US
dc.contributor.schoolSchool of Computer Engineeringen_US
dc.contributor.supervisorOng Yew Soonen_US
dc.description.degreeDOCTOR OF PHILOSOPHY (SCE)en_US

Files in this item


This item appears in the following Collection(s)

Show simple item record