Wavelet-based self-organizing maps for classifying multivariate time series

Pierpaolo D'Urso, Livia De Giovanni, Elizabeth Ann Maharaj, Riccardo Massari

Research output: Contribution to journalArticleResearchpeer-review

Abstract

Following a nonparametric approach, we suggest a time-series clustering method. Our clustering approach combines the benefits connected to the interpretative power of the nonparametric representation of the time series, and the clustering and vector quantization informational gain produced by the adopted unsupervised neural networks technique, enhanced with the self-organizing maps ordering and topological preservation abilities. The proposed clustering method takes into account a composite wavelet-based information of the multivariate time series by adding to the information connected to the wavelet variance, namely the influence of variability of individual univariate components of them ultivariate time series across scales, the information associated to wavelet correlation, represented by the interaction between pairs of univariate components of the multivariate time series at each scale, and then suitably tuning the combination of these pieces of information. In order to assess the effectiveness of the proposed clustering approach, a simulation study and an empirical application are shown.
Original languageEnglish
Pages (from-to)28 - 51
Number of pages24
JournalJournal of Chemometrics
Volume28
DOIs
Publication statusPublished - 2014

Cite this

D'Urso, Pierpaolo ; De Giovanni, Livia ; Maharaj, Elizabeth Ann ; Massari, Riccardo. / Wavelet-based self-organizing maps for classifying multivariate time series. In: Journal of Chemometrics. 2014 ; Vol. 28. pp. 28 - 51.
@article{1bd07ddf595841a5b0169959c6cf1ade,
title = "Wavelet-based self-organizing maps for classifying multivariate time series",
abstract = "Following a nonparametric approach, we suggest a time-series clustering method. Our clustering approach combines the benefits connected to the interpretative power of the nonparametric representation of the time series, and the clustering and vector quantization informational gain produced by the adopted unsupervised neural networks technique, enhanced with the self-organizing maps ordering and topological preservation abilities. The proposed clustering method takes into account a composite wavelet-based information of the multivariate time series by adding to the information connected to the wavelet variance, namely the influence of variability of individual univariate components of them ultivariate time series across scales, the information associated to wavelet correlation, represented by the interaction between pairs of univariate components of the multivariate time series at each scale, and then suitably tuning the combination of these pieces of information. In order to assess the effectiveness of the proposed clustering approach, a simulation study and an empirical application are shown.",
author = "Pierpaolo D'Urso and {De Giovanni}, Livia and Maharaj, {Elizabeth Ann} and Riccardo Massari",
year = "2014",
doi = "10.1002/cem.2565",
language = "English",
volume = "28",
pages = "28 -- 51",
journal = "Journal of Chemometrics",
issn = "0886-9383",
publisher = "Wiley-Blackwell",

}

Wavelet-based self-organizing maps for classifying multivariate time series. / D'Urso, Pierpaolo; De Giovanni, Livia; Maharaj, Elizabeth Ann; Massari, Riccardo.

In: Journal of Chemometrics, Vol. 28, 2014, p. 28 - 51.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - Wavelet-based self-organizing maps for classifying multivariate time series

AU - D'Urso, Pierpaolo

AU - De Giovanni, Livia

AU - Maharaj, Elizabeth Ann

AU - Massari, Riccardo

PY - 2014

Y1 - 2014

N2 - Following a nonparametric approach, we suggest a time-series clustering method. Our clustering approach combines the benefits connected to the interpretative power of the nonparametric representation of the time series, and the clustering and vector quantization informational gain produced by the adopted unsupervised neural networks technique, enhanced with the self-organizing maps ordering and topological preservation abilities. The proposed clustering method takes into account a composite wavelet-based information of the multivariate time series by adding to the information connected to the wavelet variance, namely the influence of variability of individual univariate components of them ultivariate time series across scales, the information associated to wavelet correlation, represented by the interaction between pairs of univariate components of the multivariate time series at each scale, and then suitably tuning the combination of these pieces of information. In order to assess the effectiveness of the proposed clustering approach, a simulation study and an empirical application are shown.

AB - Following a nonparametric approach, we suggest a time-series clustering method. Our clustering approach combines the benefits connected to the interpretative power of the nonparametric representation of the time series, and the clustering and vector quantization informational gain produced by the adopted unsupervised neural networks technique, enhanced with the self-organizing maps ordering and topological preservation abilities. The proposed clustering method takes into account a composite wavelet-based information of the multivariate time series by adding to the information connected to the wavelet variance, namely the influence of variability of individual univariate components of them ultivariate time series across scales, the information associated to wavelet correlation, represented by the interaction between pairs of univariate components of the multivariate time series at each scale, and then suitably tuning the combination of these pieces of information. In order to assess the effectiveness of the proposed clustering approach, a simulation study and an empirical application are shown.

U2 - 10.1002/cem.2565

DO - 10.1002/cem.2565

M3 - Article

VL - 28

SP - 28

EP - 51

JO - Journal of Chemometrics

JF - Journal of Chemometrics

SN - 0886-9383

ER -