摘要
为克服区间集决策表现有不确定性度量方法的不足,首先提出δ-区间改进近似粗糙度,然后将其变形与δ-条件信息熵求和,提出一种修正的δ-区间决策条件熵方法.与现有的方法相比,它同时考虑了由上、下近似集和粒化结构引起的不确定性,从理论上证明了集成后的不确定性度量值大于单个影响因素产生的不确定性度量值.最后采用区间集决策表实例进行有效验证,结果表明所提度量优于现有不确定性度量,是一种更为合理的度量方法.
In order to overcome the shortcomings of the existing uncertainty measurement methods for interval set decision table,anδ-interval improved approximate roughness is proposed,and then the deformation andδ-conditional information entropy are summed up,and thus a modifiedδ-interval decision conditional entropy method is worked out.Compared with the existing methods,the new method takes also into consideration the uncertainty caused by the upper and lower approximation sets and the granulation structure.It is theoretically proved that the value of the integrated uncertainty measurement is indeed greater than the value of the uncertainty measurement produced under a single influencing factor.Finally,an example of an interval set decision table is used for validity test,whose results show that the proposed measurement is superior to the existing uncertainty measurement,and is thus a more reasonable measurement method.
作者
唐鹏飞
TANG Pengfei(School of Mathematical Sciences, Sichuan Normal University, Chengdu 610066, China)
出处
《内江师范学院学报》
CAS
2021年第6期34-39,共6页
Journal of Neijiang Normal University
关键词
粗糙集
不确定性度量
区间集决策表
修正δ-区间决策条件熵
rough set
uncertainty measurement
interval set decision table
modifiedδ-interval decision conditional entropy
