摘要
基于粗糙集理论,改进了C D分辨矩阵,并设计了计算条件核属性的方法。通过分析指出了第二种条件核属性的计算方法,可以避免求取C D分辨矩阵,降低程序的复杂度。以条件核属性为基础,提出一种适用于获取模糊规则的数据约简算法,并通过仿真研究说明了该数据约简、模糊规则挖掘方法的有效性。
According to rough set theory, this paper improves C-D discernibility matrices and designs a program to compute condition attribute core based on the improved C-D discernibility matrices. After analysis, another program is proposed to avoid direct calculation of discerniblity matrices and improve program speed. A new data reduction algorithm of fuzzy rules extraction is presented employing condition attribute core. The simulation research shows this algorithm is effective to reduce data and extract fuzzy rules.
出处
《太原理工大学学报》
CAS
2004年第5期517-519,共3页
Journal of Taiyuan University of Technology
基金
国家自然科学基金资助项目(60374029)
山西省教育厅山西高校科技研究开发项目(2003204)
山西省青年科学基金资助项目(20041015)
