特征函数描述下的多粒度覆盖粗糙集及其不确定性度量  

Covering-based Multigranulation Rough Sets and Their Uncertainty Measures under the Description of Characteristic Function

作　　者：刘财辉[1] 凌敏 钱进 LIU Caihui;LING Min;QIAN Jin(School of Mathematics and Computer Science,Gannan Normal University,Ganzhou,Jiangxi 341000,China;School of Software,East China Jiaotong University,Nanchang 330013,China)

机构地区：[1]赣南师范大学数学与计算机科学学院,江西赣州341000 [2]华东交通大学软件学院,江西南昌330013

出　　处：《昆明理工大学学报（自然科学版）》2022年第3期51-59,137,共10页Journal of Kunming University of Science and Technology(Natural Science)

基　　金：国家自然科学基金项目(62166001,62066014,61663002);江西省自然科学基金项目(20202BAB202010,20202BABL202018);江西省“双千计划”项目;江苏省“青蓝工程”中青年学术带头人培养项目。

摘　　要：多粒度粗糙集模型是粒计算研究领域的一个热点.本文在覆盖粗糙集模型最小描述交的基础上定义了支持特征函数,设计了广义、悲观和乐观三种新的多粒度覆盖粗糙集模型,并探讨了这三种模型的相关性质.同时,分别从上、下近似及边界域的角度出发,提出了两类度量多粒度覆盖粗糙集不确定性的计算方法,并讨论了在阈值变化情况下不同度量的变化情况.最后,此次研究在4个UCI数据集上进行了相关实验,实验结果验证了方法的有效性和可行性.The multi-granulation rough sets model is one of the important issues in the field of rough set.In this paper,the supporting characteristic function is defined on the basis of the minimum description intersection in the covering rough set model.Furthermore,the generalized,pessimistic and optimistic multi-granulation covering rough sets are designed and the corresponding properties of the three models are discussed.From the perspectives of the upper and lower approximations and boundary domains,two methods of measuring the uncertainty of multi-granulation covering rough sets are proposed,and the changes of the metric are discussed when the threshold changes.Finally,this study conducted related experiments on four UCI data sets,and the experimental results verified the effectiveness and feasibility of the method.

关 键 词：多粒度覆盖粗糙集 不确定性度量 最小描述 支持特征函数 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]