Journal of Classification

, Volume 31, Issue 3, pp 325–350

Robust Functional Supervised Classification for Time Series

  • Andrés M. Alonso
  • David Casado
  • Sara López-Pintado
  • Juan Romo
Article

DOI: 10.1007/s00357-014-9163-x

Cite this article as:
Alonso, A.M., Casado, D., López-Pintado, S. et al. J Classif (2014) 31: 325. doi:10.1007/s00357-014-9163-x

Abstract

We propose using the integrated periodogram to classify time series. The method assigns a new time series to the group that minimizes the distance between the series integrated periodogram and the group mean of integrated periodograms. Local computation of these periodograms allows the application of this approach to nonstationary time series. Since the integrated periodograms are curves, we apply functional data depth-based techniques to make the classification robust, which is a clear advantage over other competitive procedures. The method provides small error rates for both simulated and real data. It improves existing approaches and presents good computational behavior.

Keywords

Time series Supervised classification Integrated periodogram Functional data depth 

Copyright information

© Classification Society of North America 2014

Authors and Affiliations

  • Andrés M. Alonso
    • 1
    • 2
  • David Casado
    • 3
  • Sara López-Pintado
    • 4
  • Juan Romo
    • 1
  1. 1.Universidad Carlos III de MadridMadridSpain
  2. 2.INAECUMadridSpain
  3. 3.Universidad Complutense de MadridMadridSpain
  4. 4.Columbia UniversityNew YorkUSA

Personalised recommendations