期刊名称：Advances in Computer Science and its Applications
印刷版ISSN：2166-2924
出版年度：2012
卷号：1
期号：3
页码：201-205
语种：English
出版社：World Science Publisher
摘要：The global optimization is the best choice for parameter extraction of rule-based classifier. Restricted methods have been published, and their limitations are concerned mainly on the slow convergence and being trapped into local minima. To resolve the matter, this paper introduced in the fitness scaling genetic algorithm (FSGA) which conducted the heuristic search as the parameter optimization for rule-based classifier. The FSGA rule-based classifier was compared with GA, SA, and ACA, and the results prove that the proposed FSGA rule-based classifier is the most robust and rapid.
