On Some Fuzzy Clustering Algorithms for Time-Series Data

Mizuki Fujita, Yuchi Kanzawa

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Various fuzzy clustering algorithms have been proposed for vectorial data. However, these methods have not been applied to time-series data. This paper presents three fuzzy clustering algorithms for time-series data based on dynamic time warping (DTW). The first algorithm involves Kullback–Leibler divergence regularization of the DTW k-means objective function. The second algorithm replaces the membership of the DTW k-means objective function with its power. The third algorithm involves q-divergence regularization of the objective function of the first algorithm. Theoretical discussion shows that the third algorithm is a generalization of the first and second algorithms, which is substantiated through numerical experiments.

Original languageEnglish
Title of host publicationIntegrated Uncertainty in Knowledge Modelling and Decision Making - 9th International Symposium, IUKM 2022, Proceedings
EditorsKatsuhiro Honda, Tomoe Entani, Seiki Ubukata, Van-Nam Huynh, Masahiro Inuiguchi
PublisherSpringer Science and Business Media Deutschland GmbH
Pages169-181
Number of pages13
ISBN (Print)9783030980177
DOIs
Publication statusPublished - 2022
Event9th International Symposium on Integrated Uncertainty in Knowledge Modelling and Decision Making, IUKM 2022 - Ishikawa, Japan
Duration: 2022 Mar 182022 Mar 19

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume13199 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference9th International Symposium on Integrated Uncertainty in Knowledge Modelling and Decision Making, IUKM 2022
Country/TerritoryJapan
CityIshikawa
Period22/3/1822/3/19

Keywords

  • Dynamic time warping
  • Fuzzy clustering
  • Time-series data

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint

Dive into the research topics of 'On Some Fuzzy Clustering Algorithms for Time-Series Data'. Together they form a unique fingerprint.

Cite this