基于条件信息熵的区间集决策信息表不确定性度量  

Uncertainty measurement for interval set decision information tables based on conditional information entropy

作　　者：张倚萌 贾修一 唐振民 Zhang Yimeng;Jia Xiuyi;Tang Zhenmin(School of Computer Science and Engineering,Nanjing University of Science andTechnology,Nanjing 210094,China)

机构地区：[1]南京理工大学计算机科学与工程学院

出　　处：《南京理工大学学报》2019年第4期393-401,共9页Journal of Nanjing University of Science and Technology

基　　金：国家自然科学基金(61773208;71671086)

摘　　要：该文研究区间集决策信息表中基于信息熵的不确定性度量。针对区间集决策信息表,该文提出一个δ-区间相似关系来描述对象之间的关系。将Pawlak粗糙集模型的近似精度和近似粗糙度等不确定性度量概念,扩展到区间集决策信息表中。通过分析扩展的δ-区间近似粗糙度和δ-区间近似精度,可以发现这两种度量对粒度结构的变化并不敏感。结合条件信息熵,该文提出了一种δ-区间决策条件熵来度量区间集决策信息表的不确定性。对δ-区间近似粗糙度,δ-区间近似精度和δ-区间决策条件熵相关性质进行了分析和证明。通过实例验证了δ-区间决策条件熵能够有效、准确地度量区间集决策信息表的不确定性。This paper aims at studying uncertainty measures for interval set decision information tables based on conditional information entropy.A binary similarity relation,calledδ-interval similarity relation in an interval set decision table is proposed to depict the relationships of objects.Based on this relation,the extended uncertainty measures from Pawlak rough set model,namely,approximate accuracy and approximate roughness,are defined in interval set decision information tables.According to the analysis ofδ-interval approximate accuracy andδ-interval approximate roughness,they are not sensitive to the variation of granular structure.A new uncertainty measure calledδ-interval decision conditional entropy is proposed by combining with conditional information entropy in interval set decision tables.The associated properties ofδ-interval approximate accuracy,δ-interval approximate roughness andδ-interval decision conditional entropy are analyzed and proved.Through an actual example,the proposedδ-interval decision conditional entropy can measure the uncertainty of interval set decision tables effectively and accurately.

关 键 词：粗糙集理论 不确定性度量 区间集决策表 近似集合 粒计算 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]