Fuzzy segmentation of multivariate time series with KPCA and G-G clustering

被引：0
|
作者：
Wang L. [1 ,2 ]
Zhu H. [1 ,2 ]
机构：
[1] School of Automation and Electrical Engineering, University of Science and Technology Beijing, Beijing
[2] Key Laboratory of Knowledge Automation for Industrial Processes of Ministry of Education, University of Science and Technology Beijing, Beijing
来源：
Kongzhi yu Juece/Control and Decision | 2021年 / 36卷 / 01期
关键词：
Clustering; Feature extraction; Fuzzy segmentation; MDBI index; Multivariate time series;
D O I：
10.13195/j.kzyjc.2019.0849
中图分类号：
学科分类号：
摘要：
The traditional Gath-Geva (G-G) fuzzy segmentation algorithm needs to set parameters and has low segmentation efficiency for high-dimensional time series. To address such matters, a fuzzy segmentation method for multivariate time series based on kernel principal component analysis (KPCA) and G-G clustering is proposed. Firstly, the KPCA is used to extract key features of multivariate time series to remove the impacts of redundant and irrelevant variables. Then, the upper bound for the number of segments is determined using the affinity propagation (AP). Finally, taking the time information into account, the segmentation of high-dimension multivariate time series in the framework of low-dimension time series is realized with the modified Davies-Bouldin index (MDBI) and G-G fuzzy clustering. The experimental results show that the proposed algorithm can detect some sudden and gradual change trend of time series quickly and effectively, which improves the accuracy and operation efficiency. Copyright ©2021 Control and Decision.
引用
收藏
页码：115 / 124
页数：9
相关论文
共 25 条
  • [1] Esling P, Agon C., Time-series data mining, ACM Computing Surveys, 45, 1, pp. 1-34, (2012)
  • [2] Chamroukhi F, Mohammed S, Trabelsi D, Et al., Joint segmentation of multivariate time series with hidden process regression for human activity recognition, Neurocomputing, 120, pp. 633-644, (2013)
  • [3] Bar-Joseph Z, Gitter A, Simon I., Studying and modelling dynamic biological processes using time-series gene expression data, Nature Reviews Genetics, 13, 8, pp. 552-564, (2012)
  • [4] Nystrup P, Hansen B W, Larsen H O, Et al., Dynamic allocation or diversification: A regime-based approach to multiple assets, The Journal of Portfolio Management, 44, 2, pp. 62-73, (2017)
  • [5] Dobos L, Abonyi J., On-line detection of homogeneous operation ranges by dynamic principal component analysis based time-series segmentation, Chemical Engineering Science, 75, pp. 96-105, (2012)
  • [6] Fu T C., A review on time series data mining, Engineering Applications of Artificial Intelligence, 24, 1, pp. 164-181, (2011)
  • [7] Gedikli A, Aksoy H, Unal N E., Segmentation algorithm for long time series analysis, Stochastic Environmental Research & Risk Assessment, 22, 3, pp. 291-302, (2008)
  • [8] Kehagias A, Nidelkou E, Petridis V., A dynamic programming segmentation procedure for hydrological and environmental time series, Stochastic Environmental Research and Risk Assessment, 20, 1, pp. 77-94, (2006)
  • [9] Gedikli A, Aksoy H, Unal N E, Et al., Modified dynamic programming approach for offline segmentation of long hydrometeorological time series, Stochastic Environmental Research and Risk Assessment, 24, 5, pp. 547-557, (2010)
  • [10] Kehagias A., A hidden Markov model segmentation procedure for hydrological and environmental time series, Stochastic Environmental Research and Risk Assessment, 18, 2, pp. 117-130, (2004)
123