摘要
为解决粗糙集属性约简算法存在的诸多问题,从属性相似度出发推导出属性相似度与粒度相似度的一致性,进而提出了一种基于属性相似度的属性约简算法,主要包括采用分明矩阵法求条件属性集的约简,核的求取,可省属性的相似度计算和最简约简的求出等步骤.仿真与对比分析表明其计算简便、效果显著.
There are many problems in the attribute reduction algorithm of rough set, in order to solve these problems, the consistency between attribute similarity and granularity similarity is deduced, moreover, an algorithm of attribute reduction is presented base on attribute similarity, the main steps include acquiring the condition attribute reduction by the indiscernible matrix method, determining the core, calculating the similarity of dispensable attribute and searching for the most simple reduced attribution. The simulation and comparative analysis show the calculation is simple and the effect is notable.
出处
《河北工业大学学报》
CAS
2005年第4期20-23,共4页
Journal of Hebei University of Technology
基金
国家自然科学基金资助项目(60173058
60377020)
关键词
属性约简
属性相似度
粒度
attribute reduction
attribute similarity
granularity
