基于一致规则的知识约简方法  

Method of knowledge reduction based on consistent rules

作　　者：李永森[1] 杨善林[1] 马溪骏[1] 秦科[1] 

机构地区：[1]合肥工业大学管理学院,安徽合肥230009

出　　处：《合肥工业大学学报（自然科学版）》2005年第10期1276-1280,共5页Journal of Hefei University of Technology：Natural Science

摘　　要：粗糙集理论应用于决策表知识约简的主要思想是在保持信息系统分类能力不变的前提下,通过寻找最佳知识约简导出问题的决策和分类规则。传统方法多是基于知识分类的相对知识约简研究,算法较为复杂;文章从决策规则本身出发,依据规则的一致性进行算法设计,方便快捷地实现了决策表的属性约简和属性值约简。The main idea of the theory of rough sets is to find out decision making and classification rules through knowledge reduction without changing the classification capacity of the information system. Traditional measures are relative knowledge reduction based on knowledge classification, which makes it difficult to find out the reduction. This study is based on decision making rules themselves, and consistent rules are used, which makes it very effective and convenient to achieve not only consistent and inconsistent decision table reduction but also attribute and value reduction.

关 键 词：粗糙集 一致规则 决策表 知识约简 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]