A fast incremental spectral clustering for large data sets

Tengteng Kong, Ye Tian, Hong Shen

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

16 Citations (Scopus)

Abstract

Spectral clustering is an emerging research topic that has numerous applications, such as data dimension reduction and image segmentation. In spectral clustering, as new data points are added continuously, dynamic data sets are processed in an on-line way to avoid costly re-computation. In this paper, we propose a new representative measure to compress the original data sets and maintain a set of representative points by continuously updating Eigen-system with the incidence vector. According to these extracted points we generate instant cluster labels as new data points arrive. Our method is effective and able to process large data sets due to its low time complexity. Experimental results over various real evolutional data sets show that our method provides fast and relatively accurate results.

Original languageEnglish
Title of host publicationProceedings - 2011 12th International Conference on Parallel and Distributed Computing, Applications and Technologies, PDCAT 2011
Pages1-5
Number of pages5
DOIs
Publication statusPublished - 2011
Externally publishedYes
Event2011 12th International Conference on Parallel and Distributed Computing, Applications and Technologies, PDCAT 2011 - Gwangju, Korea, Republic of
Duration: 20 Oct 201122 Oct 2011

Publication series

NameParallel and Distributed Computing, Applications and Technologies, PDCAT Proceedings

Conference

Conference2011 12th International Conference on Parallel and Distributed Computing, Applications and Technologies, PDCAT 2011
Country/TerritoryKorea, Republic of
CityGwangju
Period20/10/1122/10/11

Keywords

  • Eigen-gap
  • Incremental
  • Representative point
  • Spectral clustering

Fingerprint

Dive into the research topics of 'A fast incremental spectral clustering for large data sets'. Together they form a unique fingerprint.

Cite this