Distributed algorithm for principal component analysis based on power method and average consensus algorithm

Norikazu Takahashi; Mutsuki Oura; Tsuyoshi Migita

doi:10.1109/PIC50277.2020.9350816

Distributed algorithm for principal component analysis based on power method and average consensus algorithm

Norikazu Takahashi, Mutsuki Oura, Tsuyoshi Migita

School of Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Principal component analysis is one of the most important methods of multivariate analysis, and has been applied in a wide range of fields such as statistical analysis, machine learning, pattern recognition, signal processing, and communication. Recently, using the idea of multi-agent networks, distributed algorithms for principal component analysis have been proposed for the case where the data matrix is partitioned either row-wise or column-wise. In this paper, considering the case where the data matrix is partitioned both row-wise and column-wise, we propose a new algorithm that allows a multi-agent network to perform principal component analysis in a distributed manner. We also verify its validity by numerical experiments. The proposed algorithm is based on the power method for principal component analysis and the average consensus algorithm.

Original language	English
Title of host publication	Proceedings of 2020 IEEE International Conference on Progress in Informatics and Computing, PIC 2020
Editors	Yinglin Wang, Yanghua Xiao
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	16-21
Number of pages	6
ISBN (Electronic)	9781728170862
DOIs	https://doi.org/10.1109/PIC50277.2020.9350816
Publication status	Published - Dec 18 2020
Event	7th IEEE International Conference on Progress in Informatics and Computing, PIC 2020 - Shanghai, China Duration: Dec 18 2020 → Dec 20 2020

Publication series

Name	Proceedings of 2020 IEEE International Conference on Progress in Informatics and Computing, PIC 2020

Conference

Conference	7th IEEE International Conference on Progress in Informatics and Computing, PIC 2020
Country/Territory	China
City	Shanghai
Period	12/18/20 → 12/20/20

Keywords

Average consensus algorithm
Power method
Principal component analysis

ASJC Scopus subject areas

Artificial Intelligence
Computer Vision and Pattern Recognition
Information Systems
Signal Processing
Control and Optimization

Access to Document

10.1109/PIC50277.2020.9350816

Cite this

Takahashi, N., Oura, M., & Migita, T. (2020). Distributed algorithm for principal component analysis based on power method and average consensus algorithm. In Y. Wang, & Y. Xiao (Eds.), Proceedings of 2020 IEEE International Conference on Progress in Informatics and Computing, PIC 2020 (pp. 16-21). Article 9350816 (Proceedings of 2020 IEEE International Conference on Progress in Informatics and Computing, PIC 2020). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/PIC50277.2020.9350816

Distributed algorithm for principal component analysis based on power method and average consensus algorithm. / Takahashi, Norikazu; Oura, Mutsuki; Migita, Tsuyoshi.
Proceedings of 2020 IEEE International Conference on Progress in Informatics and Computing, PIC 2020. ed. / Yinglin Wang; Yanghua Xiao. Institute of Electrical and Electronics Engineers Inc., 2020. p. 16-21 9350816 (Proceedings of 2020 IEEE International Conference on Progress in Informatics and Computing, PIC 2020).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Takahashi, N, Oura, M & Migita, T 2020, Distributed algorithm for principal component analysis based on power method and average consensus algorithm. in Y Wang & Y Xiao (eds), Proceedings of 2020 IEEE International Conference on Progress in Informatics and Computing, PIC 2020., 9350816, Proceedings of 2020 IEEE International Conference on Progress in Informatics and Computing, PIC 2020, Institute of Electrical and Electronics Engineers Inc., pp. 16-21, 7th IEEE International Conference on Progress in Informatics and Computing, PIC 2020, Shanghai, China, 12/18/20. https://doi.org/10.1109/PIC50277.2020.9350816

Takahashi N, Oura M, Migita T. Distributed algorithm for principal component analysis based on power method and average consensus algorithm. In Wang Y, Xiao Y, editors, Proceedings of 2020 IEEE International Conference on Progress in Informatics and Computing, PIC 2020. Institute of Electrical and Electronics Engineers Inc. 2020. p. 16-21. 9350816. (Proceedings of 2020 IEEE International Conference on Progress in Informatics and Computing, PIC 2020). doi: 10.1109/PIC50277.2020.9350816

Takahashi, Norikazu ; Oura, Mutsuki ; Migita, Tsuyoshi. / Distributed algorithm for principal component analysis based on power method and average consensus algorithm. Proceedings of 2020 IEEE International Conference on Progress in Informatics and Computing, PIC 2020. editor / Yinglin Wang ; Yanghua Xiao. Institute of Electrical and Electronics Engineers Inc., 2020. pp. 16-21 (Proceedings of 2020 IEEE International Conference on Progress in Informatics and Computing, PIC 2020).

@inproceedings{3f14b8cfafe64e599fe01d641c96d57b,

title = "Distributed algorithm for principal component analysis based on power method and average consensus algorithm",

abstract = "Principal component analysis is one of the most important methods of multivariate analysis, and has been applied in a wide range of fields such as statistical analysis, machine learning, pattern recognition, signal processing, and communication. Recently, using the idea of multi-agent networks, distributed algorithms for principal component analysis have been proposed for the case where the data matrix is partitioned either row-wise or column-wise. In this paper, considering the case where the data matrix is partitioned both row-wise and column-wise, we propose a new algorithm that allows a multi-agent network to perform principal component analysis in a distributed manner. We also verify its validity by numerical experiments. The proposed algorithm is based on the power method for principal component analysis and the average consensus algorithm.",

keywords = "Average consensus algorithm, Power method, Principal component analysis",

author = "Norikazu Takahashi and Mutsuki Oura and Tsuyoshi Migita",

note = "Publisher Copyright: {\textcopyright} 2020 IEEE.; 7th IEEE International Conference on Progress in Informatics and Computing, PIC 2020 ; Conference date: 18-12-2020 Through 20-12-2020",

year = "2020",

month = dec,

day = "18",

doi = "10.1109/PIC50277.2020.9350816",

language = "English",

series = "Proceedings of 2020 IEEE International Conference on Progress in Informatics and Computing, PIC 2020",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "16--21",

editor = "Yinglin Wang and Yanghua Xiao",

booktitle = "Proceedings of 2020 IEEE International Conference on Progress in Informatics and Computing, PIC 2020",

}

TY - GEN

T1 - Distributed algorithm for principal component analysis based on power method and average consensus algorithm

AU - Takahashi, Norikazu

AU - Oura, Mutsuki

AU - Migita, Tsuyoshi

PY - 2020/12/18

Y1 - 2020/12/18

N2 - Principal component analysis is one of the most important methods of multivariate analysis, and has been applied in a wide range of fields such as statistical analysis, machine learning, pattern recognition, signal processing, and communication. Recently, using the idea of multi-agent networks, distributed algorithms for principal component analysis have been proposed for the case where the data matrix is partitioned either row-wise or column-wise. In this paper, considering the case where the data matrix is partitioned both row-wise and column-wise, we propose a new algorithm that allows a multi-agent network to perform principal component analysis in a distributed manner. We also verify its validity by numerical experiments. The proposed algorithm is based on the power method for principal component analysis and the average consensus algorithm.

AB - Principal component analysis is one of the most important methods of multivariate analysis, and has been applied in a wide range of fields such as statistical analysis, machine learning, pattern recognition, signal processing, and communication. Recently, using the idea of multi-agent networks, distributed algorithms for principal component analysis have been proposed for the case where the data matrix is partitioned either row-wise or column-wise. In this paper, considering the case where the data matrix is partitioned both row-wise and column-wise, we propose a new algorithm that allows a multi-agent network to perform principal component analysis in a distributed manner. We also verify its validity by numerical experiments. The proposed algorithm is based on the power method for principal component analysis and the average consensus algorithm.

KW - Average consensus algorithm

KW - Power method

KW - Principal component analysis

UR - http://www.scopus.com/inward/record.url?scp=85101687142&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85101687142&partnerID=8YFLogxK

U2 - 10.1109/PIC50277.2020.9350816

DO - 10.1109/PIC50277.2020.9350816

M3 - Conference contribution

AN - SCOPUS:85101687142

T3 - Proceedings of 2020 IEEE International Conference on Progress in Informatics and Computing, PIC 2020

SP - 16

EP - 21

BT - Proceedings of 2020 IEEE International Conference on Progress in Informatics and Computing, PIC 2020

A2 - Wang, Yinglin

A2 - Xiao, Yanghua

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 7th IEEE International Conference on Progress in Informatics and Computing, PIC 2020

Y2 - 18 December 2020 through 20 December 2020

ER -

Distributed algorithm for principal component analysis based on power method and average consensus algorithm

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this