Parameter and Mixture Component Estimation in Spatial Hidden Markov Models: A Comparative Analysis of Computational Methods

Eugene A. Opoku; Syed Ejaz Ahmed; Trisalyn Nelson; Farouk S. Nathoo

doi:10.1007/978-3-030-49829-0_25

Parameter and Mixture Component Estimation in Spatial Hidden Markov Models: A Comparative Analysis of Computational Methods

Eugene A. Opoku, Syed Ejaz Ahmed, Trisalyn Nelson, Farouk S. Nathoo

Geographical Sciences and Urban Planning, School of (SGSUP)

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

2 Scopus citations

Abstract

Hidden Markov models incorporating the Potts model for the labelling process are an important class of mixture models in spatial statistics. These models have been applied to problems in statistical mechanics, image analysis and disease mapping, among other areas. Jointly estimating the model parameters, the discrete state variables and the number of states (number of mixture components) is recognized as a difficult combinatorial optimization problem. We make comparisons between iterated conditional modes (ICM), simulated annealing (SA) and hybrid ICM with ant colony system (ACS-ICM) optimization for pixel labelling, parameter estimation and mixture component estimation. These comparisons are made for different levels of spatial dependence in the underlying true image. Our studies demonstrate that estimation based on ACS-ICM when carefully tuned exhibits performance that is uniformly superior to both ICM as well as a carefully tuned SA algorithm.

Original language	English (US)
Title of host publication	Proceedings of the 14th International Conference on Management Science and Engineering Management, ICMSEM 2020 - Volume 1
Editors	Jiuping Xu, Gheorghe Duca, Syed Ejaz Ahmed, Fausto Pedro García Márquez, Asaf Hajiyev
Publisher	Springer
Pages	340-355
Number of pages	16
ISBN (Print)	9783030498283
DOIs	https://doi.org/10.1007/978-3-030-49829-0_25
State	Published - 2020
Event	14th International Conference on Management Science and Engineering Management, ICMSEM 2020 - Chisinau, Moldova, Republic of Duration: Jul 30 2020 → Aug 2 2020

Publication series

Name	Advances in Intelligent Systems and Computing
Volume	1190 AISC
ISSN (Print)	2194-5357
ISSN (Electronic)	2194-5365

Conference

Conference	14th International Conference on Management Science and Engineering Management, ICMSEM 2020
Country/Territory	Moldova, Republic of
City	Chisinau
Period	7/30/20 → 8/2/20

Keywords

Ant colony system optimization
Hidden Markov random field
Iterated conditional modes
Mixture model
Potts model
Pseudo-likelihood
Simulated annealing

ASJC Scopus subject areas

Control and Systems Engineering
General Computer Science

Access to Document

10.1007/978-3-030-49829-0_25

Cite this

Opoku, E. A., Ahmed, S. E., Nelson, T., & Nathoo, F. S. (2020). Parameter and Mixture Component Estimation in Spatial Hidden Markov Models: A Comparative Analysis of Computational Methods. In J. Xu, G. Duca, S. E. Ahmed, F. P. García Márquez, & A. Hajiyev (Eds.), Proceedings of the 14th International Conference on Management Science and Engineering Management, ICMSEM 2020 - Volume 1 (pp. 340-355). (Advances in Intelligent Systems and Computing; Vol. 1190 AISC). Springer. https://doi.org/10.1007/978-3-030-49829-0_25

Parameter and Mixture Component Estimation in Spatial Hidden Markov Models: A Comparative Analysis of Computational Methods. / Opoku, Eugene A.; Ahmed, Syed Ejaz; Nelson, Trisalyn et al.
Proceedings of the 14th International Conference on Management Science and Engineering Management, ICMSEM 2020 - Volume 1. ed. / Jiuping Xu; Gheorghe Duca; Syed Ejaz Ahmed; Fausto Pedro García Márquez; Asaf Hajiyev. Springer, 2020. p. 340-355 (Advances in Intelligent Systems and Computing; Vol. 1190 AISC).

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

Opoku, EA, Ahmed, SE, Nelson, T & Nathoo, FS 2020, Parameter and Mixture Component Estimation in Spatial Hidden Markov Models: A Comparative Analysis of Computational Methods. in J Xu, G Duca, SE Ahmed, FP García Márquez & A Hajiyev (eds), Proceedings of the 14th International Conference on Management Science and Engineering Management, ICMSEM 2020 - Volume 1. Advances in Intelligent Systems and Computing, vol. 1190 AISC, Springer, pp. 340-355, 14th International Conference on Management Science and Engineering Management, ICMSEM 2020, Chisinau, Moldova, Republic of, 7/30/20. https://doi.org/10.1007/978-3-030-49829-0_25

Opoku EA, Ahmed SE, Nelson T, Nathoo FS. Parameter and Mixture Component Estimation in Spatial Hidden Markov Models: A Comparative Analysis of Computational Methods. In Xu J, Duca G, Ahmed SE, García Márquez FP, Hajiyev A, editors, Proceedings of the 14th International Conference on Management Science and Engineering Management, ICMSEM 2020 - Volume 1. Springer. 2020. p. 340-355. (Advances in Intelligent Systems and Computing). doi: 10.1007/978-3-030-49829-0_25

Opoku, Eugene A. ; Ahmed, Syed Ejaz ; Nelson, Trisalyn et al. / Parameter and Mixture Component Estimation in Spatial Hidden Markov Models : A Comparative Analysis of Computational Methods. Proceedings of the 14th International Conference on Management Science and Engineering Management, ICMSEM 2020 - Volume 1. editor / Jiuping Xu ; Gheorghe Duca ; Syed Ejaz Ahmed ; Fausto Pedro García Márquez ; Asaf Hajiyev. Springer, 2020. pp. 340-355 (Advances in Intelligent Systems and Computing).

@inproceedings{60cf7126da1947e087e86244f79c6cd1,

title = "Parameter and Mixture Component Estimation in Spatial Hidden Markov Models: A Comparative Analysis of Computational Methods",

abstract = "Hidden Markov models incorporating the Potts model for the labelling process are an important class of mixture models in spatial statistics. These models have been applied to problems in statistical mechanics, image analysis and disease mapping, among other areas. Jointly estimating the model parameters, the discrete state variables and the number of states (number of mixture components) is recognized as a difficult combinatorial optimization problem. We make comparisons between iterated conditional modes (ICM), simulated annealing (SA) and hybrid ICM with ant colony system (ACS-ICM) optimization for pixel labelling, parameter estimation and mixture component estimation. These comparisons are made for different levels of spatial dependence in the underlying true image. Our studies demonstrate that estimation based on ACS-ICM when carefully tuned exhibits performance that is uniformly superior to both ICM as well as a carefully tuned SA algorithm.",

keywords = "Ant colony system optimization, Hidden Markov random field, Iterated conditional modes, Mixture model, Potts model, Pseudo-likelihood, Simulated annealing",

author = "Opoku, {Eugene A.} and Ahmed, {Syed Ejaz} and Trisalyn Nelson and Nathoo, {Farouk S.}",

note = "Funding Information: Research is supported NSERC, the Visual and Automated Disease Analytics (VADA) graduate training program and CANSSI. Publisher Copyright: {\textcopyright} 2020, The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG.; 14th International Conference on Management Science and Engineering Management, ICMSEM 2020 ; Conference date: 30-07-2020 Through 02-08-2020",

year = "2020",

doi = "10.1007/978-3-030-49829-0_25",

language = "English (US)",

isbn = "9783030498283",

series = "Advances in Intelligent Systems and Computing",

publisher = "Springer",

pages = "340--355",

editor = "Jiuping Xu and Gheorghe Duca and Ahmed, {Syed Ejaz} and {Garc{\'i}a M{\'a}rquez}, {Fausto Pedro} and Asaf Hajiyev",

booktitle = "Proceedings of the 14th International Conference on Management Science and Engineering Management, ICMSEM 2020 - Volume 1",

}

TY - GEN

T1 - Parameter and Mixture Component Estimation in Spatial Hidden Markov Models

T2 - 14th International Conference on Management Science and Engineering Management, ICMSEM 2020

AU - Opoku, Eugene A.

AU - Ahmed, Syed Ejaz

AU - Nelson, Trisalyn

AU - Nathoo, Farouk S.

N1 - Funding Information: Research is supported NSERC, the Visual and Automated Disease Analytics (VADA) graduate training program and CANSSI. Publisher Copyright: © 2020, The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG.

PY - 2020

Y1 - 2020

N2 - Hidden Markov models incorporating the Potts model for the labelling process are an important class of mixture models in spatial statistics. These models have been applied to problems in statistical mechanics, image analysis and disease mapping, among other areas. Jointly estimating the model parameters, the discrete state variables and the number of states (number of mixture components) is recognized as a difficult combinatorial optimization problem. We make comparisons between iterated conditional modes (ICM), simulated annealing (SA) and hybrid ICM with ant colony system (ACS-ICM) optimization for pixel labelling, parameter estimation and mixture component estimation. These comparisons are made for different levels of spatial dependence in the underlying true image. Our studies demonstrate that estimation based on ACS-ICM when carefully tuned exhibits performance that is uniformly superior to both ICM as well as a carefully tuned SA algorithm.

AB - Hidden Markov models incorporating the Potts model for the labelling process are an important class of mixture models in spatial statistics. These models have been applied to problems in statistical mechanics, image analysis and disease mapping, among other areas. Jointly estimating the model parameters, the discrete state variables and the number of states (number of mixture components) is recognized as a difficult combinatorial optimization problem. We make comparisons between iterated conditional modes (ICM), simulated annealing (SA) and hybrid ICM with ant colony system (ACS-ICM) optimization for pixel labelling, parameter estimation and mixture component estimation. These comparisons are made for different levels of spatial dependence in the underlying true image. Our studies demonstrate that estimation based on ACS-ICM when carefully tuned exhibits performance that is uniformly superior to both ICM as well as a carefully tuned SA algorithm.

KW - Ant colony system optimization

KW - Hidden Markov random field

KW - Iterated conditional modes

KW - Mixture model

KW - Potts model

KW - Pseudo-likelihood

KW - Simulated annealing

UR - http://www.scopus.com/inward/record.url?scp=85087765914&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85087765914&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-49829-0_25

DO - 10.1007/978-3-030-49829-0_25

M3 - Conference contribution

AN - SCOPUS:85087765914

SN - 9783030498283

T3 - Advances in Intelligent Systems and Computing

SP - 340

EP - 355

BT - Proceedings of the 14th International Conference on Management Science and Engineering Management, ICMSEM 2020 - Volume 1

A2 - Xu, Jiuping

A2 - Duca, Gheorghe

A2 - Ahmed, Syed Ejaz

A2 - García Márquez, Fausto Pedro

A2 - Hajiyev, Asaf

PB - Springer

Y2 - 30 July 2020 through 2 August 2020

ER -

Parameter and Mixture Component Estimation in Spatial Hidden Markov Models: A Comparative Analysis of Computational Methods

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this