Why Does Subsequence Time-Series Clustering Produce Sine Waves?

  • Tsuyoshi Idé
Conference paper

DOI: 10.1007/11871637_23

Part of the Lecture Notes in Computer Science book series (LNCS, volume 4213)
Cite this paper as:
Idé T. (2006) Why Does Subsequence Time-Series Clustering Produce Sine Waves?. In: Fürnkranz J., Scheffer T., Spiliopoulou M. (eds) Knowledge Discovery in Databases: PKDD 2006. PKDD 2006. Lecture Notes in Computer Science, vol 4213. Springer, Berlin, Heidelberg

Abstract

The data mining and machine learning communities were surprised when Keogh et al. (2003) pointed out that the k-means cluster centers in subsequence time-series clustering become sinusoidal pseudo-patterns for almost all kinds of input time-series data. Understanding this mechanism is an important open problem in data mining. Our new theoretical approach (based on spectral clustering and translational symmetry) explains why the cluster centers of k-means naturally tend to form sinusoidal patterns.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Tsuyoshi Idé
    • 1
  1. 1.Tokyo Research LaboratoryIBM ResearchKanagawaJapan

Personalised recommendations