摘要
文章将Rough集理论应用于不同类型的决策表(一致决策表和不一致决策表)的约简,给出了广义决策、决策规则的一致程度、属性值重要性等定义,在此基础上提出了一种基于Rough集的属性值约简算法。该算法不仅能得到更为简洁的决策规则,而且能保持决策规则的一致程度不变。实例分析表明该算法是可行的。
Rough set theory is applied to the reduction of different types of decision table(consistent decision table and inconsistent decision table).A lot of definitions such as the generalized decision,the degree of a decision rule's consistency and the significance of attribute value are given.Based on these definitions,a rough set-based algorithm for attribute value reduction is proposed.Not only can it get more concise decision rules,but also it can keep the degree of consistency unchangeable.The analysis of an example shows that the algorithm is feasible.
出处
《计算机工程与应用》
CSCD
北大核心
2003年第31期48-51,共4页
Computer Engineering and Applications
基金
国家自然科学基金资助项目(编号:60073019)
关键词
ROUGH集
决策表
值核
值约简
Rough Set,Decision Table,Core Value,Value Reduction
