基于覆盖的变精度粗糙直觉模糊集模型研究  

Research on Covering-based Variable Precision Rough Intuitionistic Fuzzy Set Models

作　　者：薛占熬[1,2] 荆萌萌 姚守倩 张艳娜 XUE Zhan′ao;JING Mengmeng;YAO Shouqian;ZHANG Yanna(College of Computer and Information Engineering, Henan Normal University, Xinxiang 453007, China;Engineering Lab of Intelligence Business & Internet of Things, Xinxiang 453007, China)

机构地区：[1]河南师范大学计算机与信息工程学院,河南新乡453007 [2]“智慧商务与物联网技术”,河南省工程实验室,河南新乡453007

出　　处：《郑州大学学报（理学版）》2022年第3期10-21,共12页Journal of Zhengzhou University:Natural Science Edition

基　　金：国家自然科学基金项目(62076089,61772176);河南省科技攻关项目(182102210078,212102210136)。

摘　　要：针对经典的Pawlak粗糙集模型容易受到噪声数据影响的问题,在覆盖概念的基础上,对变精度粗糙直觉模糊集进行研究。首先,通过设定变精度中的两个约束条件(α,β),将其引入到覆盖粗糙直觉模糊集模型中,从而提出基于覆盖的变精度粗糙直觉模糊集模型,又考虑到元素邻域、规则置信度及元素与最小描述之间的关系,定义了有关该模型的4种类型,并且证明了该模型的相关性质,分析了该模型与现有模型之间的关系以及4种模型之间的关系。其次,在所给模型的基础上定义了基于覆盖的变精度粗糙直觉模糊集模型的近似质量和粗糙性测度。最后,通过信用卡申请的实例分析证明了该模型在实际应用中的有效性,并通过改变两个约束条件(α,β)的取值,分析得出较合理的α和β取值范围。The classical Pawlak rough sets were easily affected by noise data.Based on the concept of covering,variable precision rough intuitionistic fuzzy sets were studied.Firstly,by setting two constraints in variable precision(α,β),it was introduced into covering rough intuitionistic fuzzy set models,thus,a variable precision rough intuitionistic fuzzy set model based on covering was proposed.At the same time,the neighborhood of elements,the confidence of rules,and the relationship between elements and minimum description were also considered.Four types of models were defined,the related properties of models were proved,the relationship between the model and existing models and the relationship among the four models were analyzed.Secondly,based on the given model,the approximate quality and roughness measurements of the variable precision rough intuitionistic fuzzy set models based on covering were defined.Finally,it was proved by the example of the credit card application,the effectiveness of the proposed covering variable precision rough intuitionistic fuzzy set models in the practical application was verified.And by changing the values of two constraints(α,β),a reasonable value range ofαandβwas obtained.

关 键 词：粗糙集 直觉模糊集 变精度粗糙集 覆盖粗糙模糊集 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]