SHC: a spectral algorithm for hierarchical clustering

摘要：

Hierarchical clustering (HC) is a widely used approach both in pattern recognition and data mining and has rich solutions in the literature. But all these existing solutions have some restrictions when the clustered dataset has complex structure. Spectral clustering is a graph-based, simple and outperforming method with the ability to find complex structure in dataset using spectral properties of the datasetassociated affinity matrix. In this paper, we propose a novel effective HC algorithm called SHC base on the techniques of spectral method. The experiment results both on artificial and real data sets show that our algorithm can hierarchically cluster complex data effectively and naturally.

关键词： hierarchical clustering (HC) spectral clustering eigengap

作者: Li Xiaohong Huang Jingwei

作者单位: School of Computer, Wuhan University, Wuhan, 430072 The Key Laboratory of Aerospace Information Secu School of Computer, Wuhan University, Wuhan, 430072

会议类型: 国际会议

会议名称: The First International Conference on Multimedia Information Networking and Security(第一届国际多媒体网络信息安全会议 MINES 2009)

会议地点: 武汉

会议语种:英文

页码: 876-879

在线出版日期: 2009-11-18（万方平台首次上网日期，不代表论文的发表时间）