基于病毒协同进化的分类规则提取算法  

An algorithm for classification rules extraction based on virus co-evolution theory

作　　者：马航飞[1] 

机构地区：[1]福建行政学院信息管理系,福建福州350002

出　　处：《齐齐哈尔大学学报（自然科学版）》2008年第6期12-18,共7页Journal of Qiqihar University(Natural Science Edition)

摘　　要：根据置信度和覆盖度2种准则,将类的特征明确划分为充分特征和必要特征。在此基础上结合"聚焦"的思维机制,提出了基于病毒协同进化的分类规则提取算法。使用该算法不仅可以从分类信息系统中提取出分类规则,而且更能反映人类的思维方式,有较高的分类精度。According to the confidence and coverage criteria, the characterizations of class can be divided into sufficient and necessary parts. Based on this, combined with ＂Focus＂ mode of thought,an algorithm of the virus co-evolution for classification Rules is proposed. With the extracted algorithm, it can not only extract classification rules from classification system, but also reflect more clearly the mode od human thought. This kind of algorithm has definitely better classification accuracy.

关 键 词：聚焦机制 双系统结构 遗传算法 协同进化 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]