混合信息系统中基于邻域粗糙集的双评价三支聚类算法  

Dual Evaluation Three-way Clustering Algorithm Based on Neighborhood Rough Set in the Hybrid Information System

作　　者：罗一羽 杨霁琳 张贤勇[3] 孟雄 LUO Yiyu;YANG Jilin;ZHANG Xianyong;MENG Xiong(College of Computer Science,Sichuan Normal University,Chengdu 610101,China;Visual Computing and Virtual Reality Key Laboratory of Sichuan Province(Sichuan Normal University),Chengdu 610066,China;School of Mathematical Sciences,Sichuan Normal University,Chengdu 610066,China)

机构地区：[1]四川师范大学计算机科学学院,成都610101 [2]可视化计算与虚拟现实四川省重点实验室(四川师范大学),成都610066 [3]四川师范大学数学科学学院,成都610066

出　　处：《小型微型计算机系统》2024年第10期2394-2400,共7页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(61673285)资助;教育部人文社科规划基金项目(23YJA630114)资助;四川省自然科学基金项目(2022NSFSC0929)资助;四川省科技计划项目(2022ZYD0001,2021YJ0085)资助.

摘　　要：三支聚类将不确定性高的样本置于边界域,可有效解决由数据的不确定性带来的误决策问题,因此具有良好的聚类性能.为了在混合信息系统中更合理地刻画样本间的相似性和存在的不确定性,本文提出了一种基于邻域粗糙集的双评价三支聚类算法.首先,在混合信息系统中建立广义邻域关系,并分别从样本间相似属性个数和样本间距离两个角度来建立样本间相似性的两个评价函数.然后,通过引入基于双评价函数的三支决策规则,处理了广义邻域关系下样本间的相似性和存在的不确定性.最终在混合信息系统中建立了基于广义邻域关系的双评价三支聚类模型.在UCI数据集上的实验结果证明,与已有三支聚类算法相比,本文的算法在F1-score和兰德系数上都具有较好的聚类表现.Three-way clustering places highly uncertain samples in the boundary region,effectively addressing the problem of misclassification caused by data uncertainty and demonstrating decent clustering performance.To more accurately characterize the similarity and uncertainty among samples in a hybrid information system,this study proposes a dual evaluation three-way clustering algorithm based on neighborhood rough set.Firstly,a generalized neighborhood relationship is established in the hybrid information system,and two evaluation functions for sample similarity are developed from the perspectives of the number of similar attributes and the distance between samples.Subsequently,by introducing a three-way decision rule based on the dual evaluation functions,the similarity and uncertainty between samples under the generalized neighborhood relationship are addressed.Ultimately,a dual evaluation three-way clustering model based on the generalized neighborhood relationship is established in the hybrid information system.Experimental results on the UCI dataset demonstrate that compared to existing three-way clustering algorithms,the proposed algorithm exhibits superior clustering performance in terms of F1-score and Rand index.

关 键 词：三支聚类 邻域关系 双评价函数 混合信息系统 样本相似性 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]