Forward-backward smoothing for hidden Markov models of point pattern data

Nhan Dam; Dinh Phung; Ba Ngu Vo; Viet Huynh

doi:10.1109/DSAA.2017.78

Forward-backward smoothing for hidden Markov models of point pattern data

Nhan Dam, Dinh Phung, Ba Ngu Vo, Viet Huynh

Department of Data Science & AI

Research output: Chapter in Book/Report/Conference proceeding › Conference Paper › Research › peer-review

Abstract

This paper considers a discrete-time sequential latent model for point pattern data, specifically a hidden Markov model (HMM) where each observation is an instantiation of a random finite set (RFS). This so-called RFS-HMM is worthy of investigation since point pattern data are ubiquitous in artificial intelligence and data science. We address the three basic problems typically encountered in such a sequential latent model, namely likelihood computation, hidden state inference, and parameter estimation. Moreover, we develop algorithms for solving these problems including forward-backward smoothing for likelihood computation and hidden state inference, and expectation-maximisation for parameter estimation. Simulation studies are used to demonstrate key properties of RFS-HMM, whilst real data in the domain of human dynamics are used to demonstrate its applicability.

Original language	English
Title of host publication	Proceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017
Subtitle of host publication	Tokyo, Japan 19-21 October 2017
Editors	Takashi Washio, Joao Gama , Ying Li , Rajesh Parekh , Huan Liu , Albert Bifet , Richard D. De Veaux
Place of Publication	Piscataway NJ USA
Publisher	IEEE, Institute of Electrical and Electronics Engineers
Pages	252-261
Number of pages	10
ISBN (Electronic)	9781509050048
ISBN (Print)	9781509050055
DOIs	https://doi.org/10.1109/DSAA.2017.78
Publication status	Published - 2017
Event	IEEE International Conference on Data Science and Advanced Analytics 2017 - Tokyo, Japan Duration: 19 Oct 2017 → 21 Oct 2017 Conference number: 4th http://www.dslab.it.aoyama.ac.jp/dsaa2017/ https://ieeexplore.ieee.org/xpl/conhome/8255765/proceeding (Proceedings)

Conference

Conference	IEEE International Conference on Data Science and Advanced Analytics 2017
Abbreviated title	DSAA 2017
Country/Territory	Japan
City	Tokyo
Period	19/10/17 → 21/10/17
Internet address	http://www.dslab.it.aoyama.ac.jp/dsaa2017/ https://ieeexplore.ieee.org/xpl/conhome/8255765/proceeding (Proceedings)

Keywords

Discrete states
Hidden Markov models
Point pattern data
Point processes
Random finite sets
Sequential models
Set-valued data

Access to Document

10.1109/DSAA.2017.78

Cite this

Dam, N., Phung, D., Vo, B. N., & Huynh, V. (2017). Forward-backward smoothing for hidden Markov models of point pattern data. In T. Washio, J. Gama , Y. Li , R. Parekh , H. Liu , A. Bifet , & R. D. De Veaux (Eds.), Proceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017: Tokyo, Japan 19-21 October 2017 (pp. 252-261). IEEE, Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/DSAA.2017.78

Dam, Nhan ; Phung, Dinh ; Vo, Ba Ngu et al. / Forward-backward smoothing for hidden Markov models of point pattern data. Proceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017: Tokyo, Japan 19-21 October 2017 . editor / Takashi Washio ; Joao Gama ; Ying Li ; Rajesh Parekh ; Huan Liu ; Albert Bifet ; Richard D. De Veaux . Piscataway NJ USA : IEEE, Institute of Electrical and Electronics Engineers, 2017. pp. 252-261

@inproceedings{a51837e89e024cfbaba0d92a2047e2b9,

title = "Forward-backward smoothing for hidden Markov models of point pattern data",

abstract = "This paper considers a discrete-time sequential latent model for point pattern data, specifically a hidden Markov model (HMM) where each observation is an instantiation of a random finite set (RFS). This so-called RFS-HMM is worthy of investigation since point pattern data are ubiquitous in artificial intelligence and data science. We address the three basic problems typically encountered in such a sequential latent model, namely likelihood computation, hidden state inference, and parameter estimation. Moreover, we develop algorithms for solving these problems including forward-backward smoothing for likelihood computation and hidden state inference, and expectation-maximisation for parameter estimation. Simulation studies are used to demonstrate key properties of RFS-HMM, whilst real data in the domain of human dynamics are used to demonstrate its applicability.",

keywords = "Discrete states, Hidden Markov models, Point pattern data, Point processes, Random finite sets, Sequential models, Set-valued data",

author = "Nhan Dam and Dinh Phung and Vo, {Ba Ngu} and Viet Huynh",

year = "2017",

doi = "10.1109/DSAA.2017.78",

language = "English",

isbn = "9781509050055",

pages = "252--261",

editor = "Washio, {Takashi } and {Gama }, {Joao } and {Li }, Ying and {Parekh }, Rajesh and {Liu }, {Huan } and {Bifet }, {Albert } and {D. De Veaux }, {Richard }",

booktitle = "Proceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017",

publisher = "IEEE, Institute of Electrical and Electronics Engineers",

address = "United States of America",

note = "IEEE International Conference on Data Science and Advanced Analytics 2017, DSAA 2017 ; Conference date: 19-10-2017 Through 21-10-2017",

url = "http://www.dslab.it.aoyama.ac.jp/dsaa2017/, https://ieeexplore.ieee.org/xpl/conhome/8255765/proceeding",

}

Dam, N, Phung, D, Vo, BN & Huynh, V 2017, Forward-backward smoothing for hidden Markov models of point pattern data. in T Washio, J Gama , Y Li , R Parekh , H Liu , A Bifet & R D. De Veaux (eds), Proceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017: Tokyo, Japan 19-21 October 2017 . IEEE, Institute of Electrical and Electronics Engineers, Piscataway NJ USA, pp. 252-261, IEEE International Conference on Data Science and Advanced Analytics 2017, Tokyo, Japan, 19/10/17. https://doi.org/10.1109/DSAA.2017.78

Forward-backward smoothing for hidden Markov models of point pattern data. / Dam, Nhan; Phung, Dinh; Vo, Ba Ngu et al.
Proceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017: Tokyo, Japan 19-21 October 2017 . ed. / Takashi Washio; Joao Gama ; Ying Li ; Rajesh Parekh ; Huan Liu ; Albert Bifet ; Richard D. De Veaux . Piscataway NJ USA: IEEE, Institute of Electrical and Electronics Engineers, 2017. p. 252-261.

Research output: Chapter in Book/Report/Conference proceeding › Conference Paper › Research › peer-review

TY - GEN

T1 - Forward-backward smoothing for hidden Markov models of point pattern data

AU - Dam, Nhan

AU - Phung, Dinh

AU - Vo, Ba Ngu

AU - Huynh, Viet

N1 - Conference code: 4th

PY - 2017

Y1 - 2017

N2 - This paper considers a discrete-time sequential latent model for point pattern data, specifically a hidden Markov model (HMM) where each observation is an instantiation of a random finite set (RFS). This so-called RFS-HMM is worthy of investigation since point pattern data are ubiquitous in artificial intelligence and data science. We address the three basic problems typically encountered in such a sequential latent model, namely likelihood computation, hidden state inference, and parameter estimation. Moreover, we develop algorithms for solving these problems including forward-backward smoothing for likelihood computation and hidden state inference, and expectation-maximisation for parameter estimation. Simulation studies are used to demonstrate key properties of RFS-HMM, whilst real data in the domain of human dynamics are used to demonstrate its applicability.

AB - This paper considers a discrete-time sequential latent model for point pattern data, specifically a hidden Markov model (HMM) where each observation is an instantiation of a random finite set (RFS). This so-called RFS-HMM is worthy of investigation since point pattern data are ubiquitous in artificial intelligence and data science. We address the three basic problems typically encountered in such a sequential latent model, namely likelihood computation, hidden state inference, and parameter estimation. Moreover, we develop algorithms for solving these problems including forward-backward smoothing for likelihood computation and hidden state inference, and expectation-maximisation for parameter estimation. Simulation studies are used to demonstrate key properties of RFS-HMM, whilst real data in the domain of human dynamics are used to demonstrate its applicability.

KW - Discrete states

KW - Hidden Markov models

KW - Point pattern data

KW - Point processes

KW - Random finite sets

KW - Sequential models

KW - Set-valued data

UR - http://www.scopus.com/inward/record.url?scp=85046253715&partnerID=8YFLogxK

U2 - 10.1109/DSAA.2017.78

DO - 10.1109/DSAA.2017.78

M3 - Conference Paper

SN - 9781509050055

SP - 252

EP - 261

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

A2 - Washio, Takashi

A2 - Gama , Joao

A2 - Li , Ying

A2 - Parekh , Rajesh

A2 - Liu , Huan

A2 - Bifet , Albert

A2 - D. De Veaux , Richard

PB - IEEE, Institute of Electrical and Electronics Engineers

CY - Piscataway NJ USA

T2 - IEEE International Conference on Data Science and Advanced Analytics 2017

Y2 - 19 October 2017 through 21 October 2017

ER -

Dam N, Phung D, Vo BN, Huynh V. Forward-backward smoothing for hidden Markov models of point pattern data. In Washio T, Gama J, Li Y, Parekh R, Liu H, Bifet A, D. De Veaux R, editors, Proceedings - 2017 International Conference on Data Science and Advanced Analytics, DSAA 2017: Tokyo, Japan 19-21 October 2017 . Piscataway NJ USA: IEEE, Institute of Electrical and Electronics Engineers. 2017. p. 252-261 doi: 10.1109/DSAA.2017.78

Forward-backward smoothing for hidden Markov models of point pattern data

Abstract

Conference

Keywords

Access to Document

Other files and links

Cite this