Causal patterns: Extraction of multiple causal relationships by mixture of probabilistic partial canonical correlation analysis

Hiroki Mori, Keisuke Kawano, Hiroki Yokoyama

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

1 Citation (Scopus)

Abstract

In this paper, we propose a mixture of probabilistic partial canonical correlation analysis (MPPCCA) that extracts the Causal Patterns from two multivariate time series. Causal patterns refer to the signal patterns within interactions of two elements having multiple types of mutually causal relationships, rather than a mixture of simultaneous correlations or the absence of presence of a causal relationship between the elements. In multivariate statistics, partial canonical correlation analysis (PCCA) evaluates the correlation between two multivariates after subtracting the effect of the third multivariate. PCCA can calculate the Granger Causality Index (which tests whether a time-series can be predicted from another time-series), but is not applicable to data containing multiple partial canonical correlations. After introducing the MPPCCA, we propose an expectation-maxmization (EM) algorithm that estimates the parameters and latent variables of the MPPCCA. The MPPCCA is expected to extract multiple partial canonical correlations from data series without any supervised signals to split the data as clusters. The method was then evaluated in synthetic data experiments. In the synthetic dataset, our method estimated the multiple partial canonical correlations more accurately than the existing method. To determine the types of patterns detectable by the method, experiments were also conducted on real datasets. The method estimated the communication patterns In motion-capture data. The MPPCCA is applicable to various type of signals such as brain signals, human communication and nonlinear complex multibody systems.

Original languageEnglish
Title of host publicationProceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages744-754
Number of pages11
ISBN (Electronic)9781509050048
DOIs
Publication statusPublished - 2018 Jan 16
Event4th International Conference on Data Science and Advanced Analytics, DSAA 2017 - Tokyo, Japan
Duration: 2017 Oct 192017 Oct 21

Publication series

NameProceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017
Volume2018-January

Conference

Conference4th International Conference on Data Science and Advanced Analytics, DSAA 2017
CountryJapan
CityTokyo
Period17/10/1917/10/21

Fingerprint

Partial Correlation
Canonical Correlation Analysis
Time series
Canonical Correlation
Communication
Relationships
Canonical correlation analysis
Large scale systems
Data acquisition
Brain
Multivariate Statistics
Granger Causality
Experiments
Multivariate Time Series
Statistics
Motion Capture
Multibody Systems
Latent Variables
Synthetic Data
Experiment

Keywords

  • Causal pattern
  • Granger causality
  • Mixture model
  • Probabilistic partial canonical correlation analysis

ASJC Scopus subject areas

  • Signal Processing
  • Information Systems and Management
  • Statistics, Probability and Uncertainty
  • Computer Networks and Communications

Cite this

Mori, H., Kawano, K., & Yokoyama, H. (2018). Causal patterns: Extraction of multiple causal relationships by mixture of probabilistic partial canonical correlation analysis. In Proceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017 (pp. 744-754). (Proceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017; Vol. 2018-January). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/DSAA.2017.60

Causal patterns : Extraction of multiple causal relationships by mixture of probabilistic partial canonical correlation analysis. / Mori, Hiroki; Kawano, Keisuke; Yokoyama, Hiroki.

Proceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017. Institute of Electrical and Electronics Engineers Inc., 2018. p. 744-754 (Proceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017; Vol. 2018-January).

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

Mori, H, Kawano, K & Yokoyama, H 2018, Causal patterns: Extraction of multiple causal relationships by mixture of probabilistic partial canonical correlation analysis. in Proceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017. Proceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017, vol. 2018-January, Institute of Electrical and Electronics Engineers Inc., pp. 744-754, 4th International Conference on Data Science and Advanced Analytics, DSAA 2017, Tokyo, Japan, 17/10/19. https://doi.org/10.1109/DSAA.2017.60
Mori H, Kawano K, Yokoyama H. Causal patterns: Extraction of multiple causal relationships by mixture of probabilistic partial canonical correlation analysis. In Proceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017. Institute of Electrical and Electronics Engineers Inc. 2018. p. 744-754. (Proceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017). https://doi.org/10.1109/DSAA.2017.60
Mori, Hiroki ; Kawano, Keisuke ; Yokoyama, Hiroki. / Causal patterns : Extraction of multiple causal relationships by mixture of probabilistic partial canonical correlation analysis. Proceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017. Institute of Electrical and Electronics Engineers Inc., 2018. pp. 744-754 (Proceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017).
@inproceedings{5636ff16278e4c4caf547e839c9e3e5d,
title = "Causal patterns: Extraction of multiple causal relationships by mixture of probabilistic partial canonical correlation analysis",
abstract = "In this paper, we propose a mixture of probabilistic partial canonical correlation analysis (MPPCCA) that extracts the Causal Patterns from two multivariate time series. Causal patterns refer to the signal patterns within interactions of two elements having multiple types of mutually causal relationships, rather than a mixture of simultaneous correlations or the absence of presence of a causal relationship between the elements. In multivariate statistics, partial canonical correlation analysis (PCCA) evaluates the correlation between two multivariates after subtracting the effect of the third multivariate. PCCA can calculate the Granger Causality Index (which tests whether a time-series can be predicted from another time-series), but is not applicable to data containing multiple partial canonical correlations. After introducing the MPPCCA, we propose an expectation-maxmization (EM) algorithm that estimates the parameters and latent variables of the MPPCCA. The MPPCCA is expected to extract multiple partial canonical correlations from data series without any supervised signals to split the data as clusters. The method was then evaluated in synthetic data experiments. In the synthetic dataset, our method estimated the multiple partial canonical correlations more accurately than the existing method. To determine the types of patterns detectable by the method, experiments were also conducted on real datasets. The method estimated the communication patterns In motion-capture data. The MPPCCA is applicable to various type of signals such as brain signals, human communication and nonlinear complex multibody systems.",
keywords = "Causal pattern, Granger causality, Mixture model, Probabilistic partial canonical correlation analysis",
author = "Hiroki Mori and Keisuke Kawano and Hiroki Yokoyama",
year = "2018",
month = "1",
day = "16",
doi = "10.1109/DSAA.2017.60",
language = "English",
series = "Proceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "744--754",
booktitle = "Proceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017",

}

TY - GEN

T1 - Causal patterns

T2 - Extraction of multiple causal relationships by mixture of probabilistic partial canonical correlation analysis

AU - Mori, Hiroki

AU - Kawano, Keisuke

AU - Yokoyama, Hiroki

PY - 2018/1/16

Y1 - 2018/1/16

N2 - In this paper, we propose a mixture of probabilistic partial canonical correlation analysis (MPPCCA) that extracts the Causal Patterns from two multivariate time series. Causal patterns refer to the signal patterns within interactions of two elements having multiple types of mutually causal relationships, rather than a mixture of simultaneous correlations or the absence of presence of a causal relationship between the elements. In multivariate statistics, partial canonical correlation analysis (PCCA) evaluates the correlation between two multivariates after subtracting the effect of the third multivariate. PCCA can calculate the Granger Causality Index (which tests whether a time-series can be predicted from another time-series), but is not applicable to data containing multiple partial canonical correlations. After introducing the MPPCCA, we propose an expectation-maxmization (EM) algorithm that estimates the parameters and latent variables of the MPPCCA. The MPPCCA is expected to extract multiple partial canonical correlations from data series without any supervised signals to split the data as clusters. The method was then evaluated in synthetic data experiments. In the synthetic dataset, our method estimated the multiple partial canonical correlations more accurately than the existing method. To determine the types of patterns detectable by the method, experiments were also conducted on real datasets. The method estimated the communication patterns In motion-capture data. The MPPCCA is applicable to various type of signals such as brain signals, human communication and nonlinear complex multibody systems.

AB - In this paper, we propose a mixture of probabilistic partial canonical correlation analysis (MPPCCA) that extracts the Causal Patterns from two multivariate time series. Causal patterns refer to the signal patterns within interactions of two elements having multiple types of mutually causal relationships, rather than a mixture of simultaneous correlations or the absence of presence of a causal relationship between the elements. In multivariate statistics, partial canonical correlation analysis (PCCA) evaluates the correlation between two multivariates after subtracting the effect of the third multivariate. PCCA can calculate the Granger Causality Index (which tests whether a time-series can be predicted from another time-series), but is not applicable to data containing multiple partial canonical correlations. After introducing the MPPCCA, we propose an expectation-maxmization (EM) algorithm that estimates the parameters and latent variables of the MPPCCA. The MPPCCA is expected to extract multiple partial canonical correlations from data series without any supervised signals to split the data as clusters. The method was then evaluated in synthetic data experiments. In the synthetic dataset, our method estimated the multiple partial canonical correlations more accurately than the existing method. To determine the types of patterns detectable by the method, experiments were also conducted on real datasets. The method estimated the communication patterns In motion-capture data. The MPPCCA is applicable to various type of signals such as brain signals, human communication and nonlinear complex multibody systems.

KW - Causal pattern

KW - Granger causality

KW - Mixture model

KW - Probabilistic partial canonical correlation analysis

UR - http://www.scopus.com/inward/record.url?scp=85046276055&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85046276055&partnerID=8YFLogxK

U2 - 10.1109/DSAA.2017.60

DO - 10.1109/DSAA.2017.60

M3 - Conference contribution

AN - SCOPUS:85046276055

T3 - Proceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017

SP - 744

EP - 754

BT - Proceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017

PB - Institute of Electrical and Electronics Engineers Inc.

ER -