Discrimination and clustering for multivariate time series

Yoshihide Kakizawa, Robert H. Shumway, Masanobu Taniguchi

Research output: Contribution to journalArticle

169 Citations (Scopus)

Abstract

Minimum discrimination information provides a useful generalization of likelihood methodology for classification and clustering of multivariate time series. Discrimination between different classes of multivariate time series that can be characterized by differing covariance or spectral structures is of importance in applications occurring in the analysis of geophysical and medical time series data. For discrimination between such multivariate series, Kullback-Leibler discrimination information and the Chernoff information measure are developed for the multivariate non-Gaussian case. Asymptotic error rates and limiting distributions are given for a generalized spectral disparity measure that includes the foregoing criteria as special cases. Applications to problems of clustering and classifying earthquakes and mining explosions are given.

Original languageEnglish
Pages (from-to)328-340
Number of pages13
JournalJournal of the American Statistical Association
Volume93
Issue number441
Publication statusPublished - 1998 Mar
Externally publishedYes

Fingerprint

Multivariate Time Series
Discrimination
Clustering
Information Measure
Spectral Measure
Time Series Data
Limiting Distribution
Earthquake
Explosion
Error Rate
Mining
Likelihood
Series
Multivariate time series
Methodology

Keywords

  • Chernoff
  • Divergence
  • Kullback-Leibler
  • Minimum discrimination information
  • Robustness
  • Seismology
  • Spectral analysis

ASJC Scopus subject areas

  • Mathematics(all)
  • Statistics and Probability

Cite this

Discrimination and clustering for multivariate time series. / Kakizawa, Yoshihide; Shumway, Robert H.; Taniguchi, Masanobu.

In: Journal of the American Statistical Association, Vol. 93, No. 441, 03.1998, p. 328-340.

Research output: Contribution to journalArticle

Kakizawa, Yoshihide ; Shumway, Robert H. ; Taniguchi, Masanobu. / Discrimination and clustering for multivariate time series. In: Journal of the American Statistical Association. 1998 ; Vol. 93, No. 441. pp. 328-340.
@article{82ebb2bc833f4463adab05e7532fede8,
title = "Discrimination and clustering for multivariate time series",
abstract = "Minimum discrimination information provides a useful generalization of likelihood methodology for classification and clustering of multivariate time series. Discrimination between different classes of multivariate time series that can be characterized by differing covariance or spectral structures is of importance in applications occurring in the analysis of geophysical and medical time series data. For discrimination between such multivariate series, Kullback-Leibler discrimination information and the Chernoff information measure are developed for the multivariate non-Gaussian case. Asymptotic error rates and limiting distributions are given for a generalized spectral disparity measure that includes the foregoing criteria as special cases. Applications to problems of clustering and classifying earthquakes and mining explosions are given.",
keywords = "Chernoff, Divergence, Kullback-Leibler, Minimum discrimination information, Robustness, Seismology, Spectral analysis",
author = "Yoshihide Kakizawa and Shumway, {Robert H.} and Masanobu Taniguchi",
year = "1998",
month = "3",
language = "English",
volume = "93",
pages = "328--340",
journal = "Journal of the American Statistical Association",
issn = "0162-1459",
publisher = "Taylor and Francis Ltd.",
number = "441",

}

TY - JOUR

T1 - Discrimination and clustering for multivariate time series

AU - Kakizawa, Yoshihide

AU - Shumway, Robert H.

AU - Taniguchi, Masanobu

PY - 1998/3

Y1 - 1998/3

N2 - Minimum discrimination information provides a useful generalization of likelihood methodology for classification and clustering of multivariate time series. Discrimination between different classes of multivariate time series that can be characterized by differing covariance or spectral structures is of importance in applications occurring in the analysis of geophysical and medical time series data. For discrimination between such multivariate series, Kullback-Leibler discrimination information and the Chernoff information measure are developed for the multivariate non-Gaussian case. Asymptotic error rates and limiting distributions are given for a generalized spectral disparity measure that includes the foregoing criteria as special cases. Applications to problems of clustering and classifying earthquakes and mining explosions are given.

AB - Minimum discrimination information provides a useful generalization of likelihood methodology for classification and clustering of multivariate time series. Discrimination between different classes of multivariate time series that can be characterized by differing covariance or spectral structures is of importance in applications occurring in the analysis of geophysical and medical time series data. For discrimination between such multivariate series, Kullback-Leibler discrimination information and the Chernoff information measure are developed for the multivariate non-Gaussian case. Asymptotic error rates and limiting distributions are given for a generalized spectral disparity measure that includes the foregoing criteria as special cases. Applications to problems of clustering and classifying earthquakes and mining explosions are given.

KW - Chernoff

KW - Divergence

KW - Kullback-Leibler

KW - Minimum discrimination information

KW - Robustness

KW - Seismology

KW - Spectral analysis

UR - http://www.scopus.com/inward/record.url?scp=0032343819&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0032343819&partnerID=8YFLogxK

M3 - Article

VL - 93

SP - 328

EP - 340

JO - Journal of the American Statistical Association

JF - Journal of the American Statistical Association

SN - 0162-1459

IS - 441

ER -