摘要
针对变精度粗糙集属性约简问题,在分析变精度粗糙集理论的基础上,分别从属性依赖度增量、互信息的增量、基于覆盖度与准确度相结合的度量以及属性的不确定性量度等角度,对属性重要度进行分析。并分别以这四个属性重要度作为启发式信息,提出变精度粗糙集属性约简的启发式算法,进而得到信息系统的最小约简,并将所给的算法应用MATLAB程序进行实现。最后,通过具体算例说明所给算法的有效性和实用性。
Heuristic algorithms of attribute reduction in variable precision rough sets （VPRS） are proposed based on the analysis of the VPRS theory. The attribute significance is studied in terms of increment of attribute dependence, increment of mutual information, measure of degree of accuracy with degree of coverage and measure of uncertainty, which are taken as heuristic information in the heuristic algorithms. Moreover, the least reduction is given. The corresponding algorithms are implemented on the MATLAB. At last, a practical example is given to show the validity and practicability of the algorithm.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2007年第12期2064-2067,共4页
Systems Engineering and Electronics
基金
国家自然科学基金资助课题(60574011)
关键词
变精度粗糙集
属性约简
属性重要度
启发式算法
variable precision rough sets
attribute reduction
attribute significance
heuristic algorithm
