Fast in-memory spectral clustering using a fixed-size approach

R. Langone, RaghvenPhDa Mall, V. Jumutc, J. A.K. Suykens

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

Spectral clustering represents a successful approach to data clustering. Despite its high performance in solving complex tasks, it is often disregarded in favor of the less accurate k-means algorithm because of its computational inefficiency. In this article we present a fast in-memory spectral clustering algorithm, which can handle millions of datapoints at a desktop PC scale. The proposed technique relies on a kernel-based formulation of the spectral clustering problem, also known as kernel spectral clustering. In particular, we use a fixed-size approach based on an approximation of the feature map via the Nyström method to solve the primal optimization problem. We experimented on several small and large scale real-world datasets to show the computational efficiency and clustering quality of the proposed algorithm.

Original languageEnglish
Title of host publicationESANN 2016 - 24th European Symposium on Artificial Neural Networks
Publisheri6doc.com publication
Pages557-562
Number of pages6
ISBN (Electronic)9782875870278
Publication statusPublished - 1 Jan 2016
Event24th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, ESANN 2016 - Bruges, Belgium
Duration: 27 Apr 201629 Apr 2016

Other

Other24th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, ESANN 2016
CountryBelgium
CityBruges
Period27/4/1629/4/16

    Fingerprint

ASJC Scopus subject areas

  • Artificial Intelligence
  • Information Systems

Cite this

Langone, R., Mall, R., Jumutc, V., & Suykens, J. A. K. (2016). Fast in-memory spectral clustering using a fixed-size approach. In ESANN 2016 - 24th European Symposium on Artificial Neural Networks (pp. 557-562). i6doc.com publication.