摘要
针对在使用粗糙集理论处理决策表时需要事先对连续属性进行离散化处理的问题,提出了一种基于连续属性分布特征的离散化算法．并通过实例分析说明该算法能在保证决策表原始分类能力不变的前提下,获得较少的断点数．
The continuous attributes in the decision table must be discretized firstly when using rough set theory. To this problem, it gives a discretization algorithm based on the distribution character of the continuous attributes in this paper. And the experiment result demonstrates that this algorithm can get less break-points when the partition ability of the original decision table keeps stable.
出处
《西南师范大学学报（自然科学版）》
CAS
CSCD
北大核心
2006年第2期107-110,共4页
Journal of Southwest China Normal University(Natural Science Edition)
基金
重庆市自然科学基金资助项目(8651)重庆工学院青年基金资助项目(2004ZQ08).
关键词
粗糙集
离散化
分布特征
rough set
discretization
distribution character
