| 一种基于二部图谱划分的聚类集成方法 |
| |
| 引用本文： | 徐森,皋军,徐秀芳,花小朋,徐静,安晶.一种基于二部图谱划分的聚类集成方法[J].控制与决策,2018,33(12):2208-2212. |
| |
| 作者姓名： | 徐森 皋军 徐秀芳 花小朋 徐静 安晶 |
| |
| 作者单位： | 盐城工学院信息工程学院，江苏盐城224001,盐城工学院信息工程学院，江苏盐城224001;江苏省媒体设计与软件技术重点实验室江南大学，江苏无锡214122,盐城工学院信息工程学院，江苏盐城224001,盐城工学院信息工程学院，江苏盐城224001,盐城工学院信息工程学院，江苏盐城224001,盐城工学院信息工程学院，江苏盐城224001 |
| |
| 基金项目： | 国家自然科学基金项目(61375001)；江苏省自然科学基金项目(BK20151299)；江苏省333工程项目；江苏省高等学校自然科学研究项目(18KJB520050)；江苏省媒体设计与软件技术重点实验室开放课题项目(18ST0201). |
| |
| 摘 要： | 将二部图模型引入聚类集成问题中,使用二部图模型同时建模对象集和超边集,充分挖掘潜藏在对象之间的相似度信息和超边提供的属性信息.设计正则化谱聚类算法解决二部图划分问题,在低维嵌入空间运行K-means++算法划分对象集,获得最终的聚类结果.在多组基准数据集上进行实验,实验结果表明所提出方法不仅能获得优越的结果,而且具有较高的运行效率.
|
| 关 键 词： | 机器学习 聚类分析 二部图模型 聚类集成 谱聚类算法 |
| 收稿时间： | 2017/7/27 0:00:00 |
| 修稿时间： | 2018/4/10 0:00:00 |
A cluster ensemble approach based on bipartite spectral graph partitioning |
| |
| XU Sen,GAO Jun,XU Xiu-fang,HUA Xiao-peng,XU Jing and AN Jing.A cluster ensemble approach based on bipartite spectral graph partitioning[J].Control and Decision,2018,33(12):2208-2212. |
| |
| Authors: | XU Sen GAO Jun XU Xiu-fang HUA Xiao-peng XU Jing and AN Jing |
| |
| Affiliation: | School of Information Engineering,Yancheng Institute of Technology,Yancheng 224001,China,School of Information Engineering,Yancheng Institute of Technology,Yancheng 224001,China;Jiangsu Key Laboratory of Media Design and Software TechnologyJiangnan University,Wuxi214122,China,School of Information Engineering,Yancheng Institute of Technology,Yancheng 224001,China,School of Information Engineering,Yancheng Institute of Technology,Yancheng 224001,China,School of Information Engineering,Yancheng Institute of Technology,Yancheng 224001,China and School of Information Engineering,Yancheng Institute of Technology,Yancheng 224001,China |
| |
| Abstract: | A bipartite graph model is brought into the cluster ensemble problem. The object set and hyperedge set are modeled simultaneously via a bipartite graph formulation considering the similarity among instances and the information provided by hyperedges collectively. A normalized spectral clustering algorithm is proposed to solve the bipartite graph partitioning problem, and the final clustering result is attained by performing K-means++ algorithm to partition object set embedded in low dimensional space. Experimental results on several baseline datasets show that the proposed approach is not only well-performed but also high efficient. |
| |
| Keywords: | |
|
| 点击此处可从《控制与决策》浏览原始摘要信息 |
|
点击此处可从《控制与决策》下载全文 |
|
