一种基于属性值重要性的启发式值约简算法

A Heuristic Value Reduction Algorithm Based on the Importance of Attribute's Value

对基于粗糙集理论的启发式值约简算法进行了研究,提出一种新算法.实例表明,新算法操作简便,执行速度较快,能有效约简数据,得到比一般算法更为简洁的规则.同时,文章采用UCI机器学习数据库中的部分数据库进行了实验测试,证明了算法的正确性和有效性.

This paper discusses the heuristic value reduction algorithm based on rough set theory. Following that it puts forward a new algorithm for value reduction. The examples show that the new algorithm operates very simple and convenient. It executes very fast and reduces the data in effect. It can reach more compact than common algorithms. At the same time,The validity and feasibility of the algorithms are demonstrated by testing some classical database in the UCI repository.