摘要
针对分类模型在构造过程中存在冗余属性的特点,提出了一种基于粗糙集理论的分类预处理技术,利用其理论的属性约简与分辨矩阵得到属性的约简集.经实例证明,此方法对提高分类模型构造效率有较好的效果.
Aimed at the characteristic of redundancy in the process of constructing classification model, this paper offers a classification reprocessing technology based on the Rough Set Theory to make use of the attribute reduction and distinguish matrix of the theory to produce the core. It is tested by the concrete example that the technology has good effect on constructing classification model.
出处
《兰州工业高等专科学校学报》
2006年第4期1-4,共4页
Journal of Lanzhou Higher Polytechnical College
基金
甘肃省自然科学基金(3ZS051-A25-047)
关键词
粗糙集
约简与核
分辨矩阵
rough set
reduction and core
distinguish matrix
