基于知识量的决策表约简算法

Decision Table Reduction Algorithm Based on Knowledge Quantity

提出决策属性集相对于条件属性集的知识量的概念,对决策表中属性的重要性进行了新的度量,并在此基础上,设计了一种基于知识量的约简算法,实例分析表明该算法能够有效得到决策表的约简。

Attribute reduction is an important problem in rough set theory. This paper presents the concept of the knowledge quantity of decision attributes with relation to given condition attributes and redefines the importance of the attribute in decision table. Then, a new reduction algorithm based on knowledge quantity is developed. A test shows that this algorithm is effective.