Stochastic Version of EM Algorithm for Nonlinear Random Change-Point Models

Hongbin Zhang; Binod Manandhar

doi:10.11159/icsta21.119

Stochastic Version of EM Algorithm for Nonlinear Random Change-Point Models

Hongbin Zhang, Binod Manandhar

Biostatistics

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

Abstract

Random effect change-point models are commonly used to infer individual-specific time of event that induces trend change of longitudinal data. Linear models are often employed before and after the change point. However, in applications such as HIV studies, a mechanistic nonlinear model can be derived for the process based on the underlying data-generation mechanisms and such nonlinear model may provide better ``predictions". In this article, we propose a random change-point model in which we model the longitudinal data by segmented nonlinear mixed effect models. Inference wise, we propose a maximum likelihood solution where we use the Stochastic Expectation-Maximization (StEM) algorithm coupled with independent multivariate rejection sampling through Gibbs’s sampler. We evaluate the method with simulations to gain insights.

Original language	English
Title of host publication	3rd International Conference on Statistics
Subtitle of host publication	Theory and Applications, ICSTA 2021
Editors	Gangaram S. Ladde, Noelle Samia
Publisher	Avestia Publishing
ISBN (Print)	9781927877913
DOIs	https://doi.org/10.11159/icsta21.119
State	Published - 2021
Event	3rd International Conference on Statistics: Theory and Applications, ICSTA 2021 - Virtual, Online Duration: Jul 29 2021 → Jul 31 2021

Publication series

Name	Proceedings of the International Conference on Statistics
ISSN (Electronic)	2562-7767

Conference

Conference	3rd International Conference on Statistics: Theory and Applications, ICSTA 2021
City	Virtual, Online
Period	7/29/21 → 7/31/21

Bibliographical note

Funding Information:
The work is supported by NIH grant R21AI147933. It is also partially supported by the City University of New York High-Performance Computing Center, College of Staten Island, funded in part by the City and State of New York, City University of New York Research Foundation and National Science Foundation grants CNS-0958379, CNS-0855217, and ACI-112611.

Publisher Copyright:
© 2021, Avestia Publishing. All rights reserved.

Keywords

Gibbs’s sampler
Multivariate rejection sampling
Nonlinear mixed effects model
Random change-point model
Stochastic version of EM

ASJC Scopus subject areas

Applied Mathematics
Computational Mathematics
Statistics and Probability
Theoretical Computer Science

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.11159/icsta21.119

Cite this

@inproceedings{524ab212ec794ac2bb4e0efdac3edd74,

title = "Stochastic Version of EM Algorithm for Nonlinear Random Change-Point Models",

abstract = "Random effect change-point models are commonly used to infer individual-specific time of event that induces trend change of longitudinal data. Linear models are often employed before and after the change point. However, in applications such as HIV studies, a mechanistic nonlinear model can be derived for the process based on the underlying data-generation mechanisms and such nonlinear model may provide better ``predictions{"}. In this article, we propose a random change-point model in which we model the longitudinal data by segmented nonlinear mixed effect models. Inference wise, we propose a maximum likelihood solution where we use the Stochastic Expectation-Maximization (StEM) algorithm coupled with independent multivariate rejection sampling through Gibbs{\textquoteright}s sampler. We evaluate the method with simulations to gain insights.",

keywords = "Gibbs{\textquoteright}s sampler, Multivariate rejection sampling, Nonlinear mixed effects model, Random change-point model, Stochastic version of EM",

author = "Hongbin Zhang and Binod Manandhar",

note = "Funding Information: The work is supported by NIH grant R21AI147933. It is also partially supported by the City University of New York High-Performance Computing Center, College of Staten Island, funded in part by the City and State of New York, City University of New York Research Foundation and National Science Foundation grants CNS-0958379, CNS-0855217, and ACI-112611. Publisher Copyright: {\textcopyright} 2021, Avestia Publishing. All rights reserved.; 3rd International Conference on Statistics: Theory and Applications, ICSTA 2021 ; Conference date: 29-07-2021 Through 31-07-2021",

year = "2021",

doi = "10.11159/icsta21.119",

language = "English",

isbn = "9781927877913",

series = "Proceedings of the International Conference on Statistics",

publisher = "Avestia Publishing",

editor = "Ladde, {Gangaram S.} and Noelle Samia",

booktitle = "3rd International Conference on Statistics",

address = "Canada",

}

Zhang, H & Manandhar, B 2021, Stochastic Version of EM Algorithm for Nonlinear Random Change-Point Models. in GS Ladde & N Samia (eds), 3rd International Conference on Statistics: Theory and Applications, ICSTA 2021., 119, Proceedings of the International Conference on Statistics, Avestia Publishing, 3rd International Conference on Statistics: Theory and Applications, ICSTA 2021, Virtual, Online, 7/29/21. https://doi.org/10.11159/icsta21.119

Stochastic Version of EM Algorithm for Nonlinear Random Change-Point Models. / Zhang, Hongbin; Manandhar, Binod.
3rd International Conference on Statistics: Theory and Applications, ICSTA 2021. ed. / Gangaram S. Ladde; Noelle Samia. Avestia Publishing, 2021. 119 (Proceedings of the International Conference on Statistics).

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

TY - GEN

T1 - Stochastic Version of EM Algorithm for Nonlinear Random Change-Point Models

AU - Zhang, Hongbin

AU - Manandhar, Binod

N1 - Funding Information: The work is supported by NIH grant R21AI147933. It is also partially supported by the City University of New York High-Performance Computing Center, College of Staten Island, funded in part by the City and State of New York, City University of New York Research Foundation and National Science Foundation grants CNS-0958379, CNS-0855217, and ACI-112611. Publisher Copyright: © 2021, Avestia Publishing. All rights reserved.

PY - 2021

Y1 - 2021

N2 - Random effect change-point models are commonly used to infer individual-specific time of event that induces trend change of longitudinal data. Linear models are often employed before and after the change point. However, in applications such as HIV studies, a mechanistic nonlinear model can be derived for the process based on the underlying data-generation mechanisms and such nonlinear model may provide better ``predictions". In this article, we propose a random change-point model in which we model the longitudinal data by segmented nonlinear mixed effect models. Inference wise, we propose a maximum likelihood solution where we use the Stochastic Expectation-Maximization (StEM) algorithm coupled with independent multivariate rejection sampling through Gibbs’s sampler. We evaluate the method with simulations to gain insights.

AB - Random effect change-point models are commonly used to infer individual-specific time of event that induces trend change of longitudinal data. Linear models are often employed before and after the change point. However, in applications such as HIV studies, a mechanistic nonlinear model can be derived for the process based on the underlying data-generation mechanisms and such nonlinear model may provide better ``predictions". In this article, we propose a random change-point model in which we model the longitudinal data by segmented nonlinear mixed effect models. Inference wise, we propose a maximum likelihood solution where we use the Stochastic Expectation-Maximization (StEM) algorithm coupled with independent multivariate rejection sampling through Gibbs’s sampler. We evaluate the method with simulations to gain insights.

KW - Gibbs’s sampler

KW - Multivariate rejection sampling

KW - Nonlinear mixed effects model

KW - Random change-point model

KW - Stochastic version of EM

UR - http://www.scopus.com/inward/record.url?scp=85161175741&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85161175741&partnerID=8YFLogxK

U2 - 10.11159/icsta21.119

DO - 10.11159/icsta21.119

M3 - Conference contribution

AN - SCOPUS:85161175741

SN - 9781927877913

T3 - Proceedings of the International Conference on Statistics

BT - 3rd International Conference on Statistics

A2 - Ladde, Gangaram S.

A2 - Samia, Noelle

PB - Avestia Publishing

T2 - 3rd International Conference on Statistics: Theory and Applications, ICSTA 2021

Y2 - 29 July 2021 through 31 July 2021

ER -

Stochastic Version of EM Algorithm for Nonlinear Random Change-Point Models

Abstract

Publication series

Conference

Bibliographical note

Keywords

ASJC Scopus subject areas

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this