摘要
在研究集值信息系统和知识距离性质的基础上,提出了一种基于知识距离的集值信息系统属性约简算法.该算法首先利用知识距离来描述知识间的差距,进而度量集值信息系统模型的划分和知识粒度的大小,然后根据知识距离的性质有效的判定集值信息系统模型的宽松度和属性重要性.最后通过理论分析和实例的结果表明,该算法降低了时空复杂度,从而提高了算法的运行效率.
A new attribute reduction algorithm is proposed for the setvalue information system based on knowledge distance on the basis of the study of setvalue information system and knowledge distance prop erties. This algorithm firstly uses knowledge distance to describe the gap between the knowledge, then measures the division of the setvalue information system and the size of the knowledge granularity, and then decides the loose degree of the setvalue information system model and the attribute importance effi ciently according to the nature of knowledge distance. At last the algorithm is proved to be viable to reduce the time and space complexity to improve the operating efficiency by the theoretical analysis and the results of examples.
出处
《兰州交通大学学报》
CAS
2013年第1期107-110,共4页
Journal of Lanzhou Jiaotong University
基金
甘肃省科技支撑计划项目(1011GKCA040)
甘肃省科技支撑计划项目(1104GKCA016)
关键词
粗糙集
集值信息系统
知识距离
属性约简
rough set
set-valued information system knowledge distance
attribute reduction
