Robust Bayesian Changepoint Analysis in the Presence of Outliers

Shonosuke Sugasawa; Shintaro Hashimoto

doi:10.1007/978-981-16-2765-1_39

Robust Bayesian Changepoint Analysis in the Presence of Outliers

Shonosuke Sugasawa, Shintaro Hashimoto

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

Abstract

We introduce a new robust Bayesian change-point analysis in the presence of outliers. We employ an idea of general posterior based on density power divergence combined with horseshoe prior for differences of underlying signals. A posterior computation algorithm is proposed using Markov chain Monte Carlo. The proposed method is demonstrated through simulation and real data analysis.

Original language	English
Title of host publication	Intelligent Decision Technologies - Proceedings of the 13th KES-IDT 2021 Conference
Editors	Ireneusz Czarnowski, Robert J. Howlett, Lakhmi C. Jain
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	469-478
Number of pages	10
ISBN (Print)	9789811627644
DOIs	https://doi.org/10.1007/978-981-16-2765-1_39
Publication status	Published - 2021
Externally published	Yes
Event	13th International KES Conference on Intelligent Decision Technologies, KES-IDT 2021 - Virtual, Online Duration: 2021 Jun 14 → 2021 Jun 16

Publication series

Name	Smart Innovation, Systems and Technologies
Volume	238
ISSN (Print)	2190-3018
ISSN (Electronic)	2190-3026

Conference

Conference	13th International KES Conference on Intelligent Decision Technologies, KES-IDT 2021
City	Virtual, Online
Period	21/6/14 → 21/6/16

Keywords

Change-point analysis
Density power divergence
Horseshoe prior
State space model

ASJC Scopus subject areas

Decision Sciences(all)
Computer Science(all)

Access to Document

10.1007/978-981-16-2765-1_39

Cite this

Sugasawa, S., & Hashimoto, S. (2021). Robust Bayesian Changepoint Analysis in the Presence of Outliers. In I. Czarnowski, R. J. Howlett, & L. C. Jain (Eds.), Intelligent Decision Technologies - Proceedings of the 13th KES-IDT 2021 Conference (pp. 469-478). (Smart Innovation, Systems and Technologies; Vol. 238). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-981-16-2765-1_39

Robust Bayesian Changepoint Analysis in the Presence of Outliers. / Sugasawa, Shonosuke; Hashimoto, Shintaro.
Intelligent Decision Technologies - Proceedings of the 13th KES-IDT 2021 Conference. ed. / Ireneusz Czarnowski; Robert J. Howlett; Lakhmi C. Jain. Springer Science and Business Media Deutschland GmbH, 2021. p. 469-478 (Smart Innovation, Systems and Technologies; Vol. 238).

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

Sugasawa, S & Hashimoto, S 2021, Robust Bayesian Changepoint Analysis in the Presence of Outliers. in I Czarnowski, RJ Howlett & LC Jain (eds), Intelligent Decision Technologies - Proceedings of the 13th KES-IDT 2021 Conference. Smart Innovation, Systems and Technologies, vol. 238, Springer Science and Business Media Deutschland GmbH, pp. 469-478, 13th International KES Conference on Intelligent Decision Technologies, KES-IDT 2021, Virtual, Online, 21/6/14. https://doi.org/10.1007/978-981-16-2765-1_39

Sugasawa S, Hashimoto S. Robust Bayesian Changepoint Analysis in the Presence of Outliers. In Czarnowski I, Howlett RJ, Jain LC, editors, Intelligent Decision Technologies - Proceedings of the 13th KES-IDT 2021 Conference. Springer Science and Business Media Deutschland GmbH. 2021. p. 469-478. (Smart Innovation, Systems and Technologies). doi: 10.1007/978-981-16-2765-1_39

Sugasawa, Shonosuke ; Hashimoto, Shintaro. / Robust Bayesian Changepoint Analysis in the Presence of Outliers. Intelligent Decision Technologies - Proceedings of the 13th KES-IDT 2021 Conference. editor / Ireneusz Czarnowski ; Robert J. Howlett ; Lakhmi C. Jain. Springer Science and Business Media Deutschland GmbH, 2021. pp. 469-478 (Smart Innovation, Systems and Technologies).

@inproceedings{6602dfff41be40dcaa4855e11f34038c,

title = "Robust Bayesian Changepoint Analysis in the Presence of Outliers",

abstract = "We introduce a new robust Bayesian change-point analysis in the presence of outliers. We employ an idea of general posterior based on density power divergence combined with horseshoe prior for differences of underlying signals. A posterior computation algorithm is proposed using Markov chain Monte Carlo. The proposed method is demonstrated through simulation and real data analysis.",

keywords = "Change-point analysis, Density power divergence, Horseshoe prior, State space model",

author = "Shonosuke Sugasawa and Shintaro Hashimoto",

note = "Funding Information: Fig. 4 Posterior medians of θt (blue lines), detected change-points (red vertical lines), and posterior probability of {θt+1 − θt ≥ 0.1} based on the HCP and RHCP methods, applied to the Well-Log data Acknowledgements The authors would like to thank the referees for the careful reading of the paper, and the valuable suggestions and comments. This work is partially supported by Japan Society for Promotion of Science (KAKENHI) grant numbers 18K12757 and 17K14233. Publisher Copyright: {\textcopyright} 2021, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.; 13th International KES Conference on Intelligent Decision Technologies, KES-IDT 2021 ; Conference date: 14-06-2021 Through 16-06-2021",

year = "2021",

doi = "10.1007/978-981-16-2765-1_39",

language = "English",

isbn = "9789811627644",

series = "Smart Innovation, Systems and Technologies",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "469--478",

editor = "Ireneusz Czarnowski and Howlett, {Robert J.} and Jain, {Lakhmi C.}",

booktitle = "Intelligent Decision Technologies - Proceedings of the 13th KES-IDT 2021 Conference",

}

TY - GEN

T1 - Robust Bayesian Changepoint Analysis in the Presence of Outliers

AU - Sugasawa, Shonosuke

AU - Hashimoto, Shintaro

N1 - Funding Information: Fig. 4 Posterior medians of θt (blue lines), detected change-points (red vertical lines), and posterior probability of {θt+1 − θt ≥ 0.1} based on the HCP and RHCP methods, applied to the Well-Log data Acknowledgements The authors would like to thank the referees for the careful reading of the paper, and the valuable suggestions and comments. This work is partially supported by Japan Society for Promotion of Science (KAKENHI) grant numbers 18K12757 and 17K14233. Publisher Copyright: © 2021, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

PY - 2021

Y1 - 2021

N2 - We introduce a new robust Bayesian change-point analysis in the presence of outliers. We employ an idea of general posterior based on density power divergence combined with horseshoe prior for differences of underlying signals. A posterior computation algorithm is proposed using Markov chain Monte Carlo. The proposed method is demonstrated through simulation and real data analysis.

AB - We introduce a new robust Bayesian change-point analysis in the presence of outliers. We employ an idea of general posterior based on density power divergence combined with horseshoe prior for differences of underlying signals. A posterior computation algorithm is proposed using Markov chain Monte Carlo. The proposed method is demonstrated through simulation and real data analysis.

KW - Change-point analysis

KW - Density power divergence

KW - Horseshoe prior

KW - State space model

UR - http://www.scopus.com/inward/record.url?scp=85112028913&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85112028913&partnerID=8YFLogxK

U2 - 10.1007/978-981-16-2765-1_39

DO - 10.1007/978-981-16-2765-1_39

M3 - Conference contribution

AN - SCOPUS:85112028913

SN - 9789811627644

T3 - Smart Innovation, Systems and Technologies

SP - 469

EP - 478

BT - Intelligent Decision Technologies - Proceedings of the 13th KES-IDT 2021 Conference

A2 - Czarnowski, Ireneusz

A2 - Howlett, Robert J.

A2 - Jain, Lakhmi C.

PB - Springer Science and Business Media Deutschland GmbH

T2 - 13th International KES Conference on Intelligent Decision Technologies, KES-IDT 2021

Y2 - 14 June 2021 through 16 June 2021

ER -

Robust Bayesian Changepoint Analysis in the Presence of Outliers

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this