Presentation 2010-03-10
ARMA Model Based Time Series Clustering Using Dirichiet Process Mixture Models
Yuki WASHIZU, Nobuo SUEMATSU, Akira HAYASHI, Kazunori IWATA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Dirichiet Process Mixture (DPM) models allow nonparametric mixture modeling in which the number of mixture components is not specified. This is quite beneficial for clustering and DPM models have been applied to many clustering applications successfully. Gibbs sampling for DPM models where conjugate priors are used is established. However, when the component models do not have conjugate priors, such as the case of ARMA models, we face some difficulty. In this paper, we cope with the difficulty by introducing underlying Metropolis-Hastings chains and realize an ARMA model based time series clustering via Dirichiet process mixture models.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Dirichlet Process Mixture Model / ARMA Model / Markov Chain Monte Carlo
Paper # NC2009-135
Date of Issue

Conference Information
Committee NC
Conference Date 2010/3/2(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Neurocomputing (NC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) ARMA Model Based Time Series Clustering Using Dirichiet Process Mixture Models
Sub Title (in English)
Keyword(1) Dirichlet Process Mixture Model
Keyword(2) ARMA Model
Keyword(3) Markov Chain Monte Carlo
1st Author's Name Yuki WASHIZU
1st Author's Affiliation Graduate School of Information Sciences, Hiroshima City University()
2nd Author's Name Nobuo SUEMATSU
2nd Author's Affiliation Graduate School of Information Sciences, Hiroshima City University
3rd Author's Name Akira HAYASHI
3rd Author's Affiliation Graduate School of Information Sciences, Hiroshima City University
4th Author's Name Kazunori IWATA
4th Author's Affiliation Graduate School of Information Sciences, Hiroshima City University
Date 2010-03-10
Paper # NC2009-135
Volume (vol) vol.109
Number (no) 461
Page pp.pp.-
#Pages 6
Date of Issue