Extreme-based clustering of environmental time series Articles uri icon

publication date

  • January 2013

start page

  • 92

end page

  • 102

issue

  • 2

volume

  • 29

international standard serial number (ISSN)

  • 1889-3805

abstract

  • This work provides an up-to-date review on clustering techniques to classify time series on the basis of their corresponding extremal properties with a bias towards describing the authors' ongoing work. Applications to clustering time series of sea-level and daily mean temperature are presented

keywords

  • extreme value theory; cluster analysis; bayesian analysis; return values