Latent Vector Autoregressive Modeling for Reduced Dimensional Dynamic Feature Extraction and Prediction

S. Joe Qin

doi:10.1109/CDC45484.2021.9683255

Latent Vector Autoregressive Modeling for Reduced Dimensional Dynamic Feature Extraction and Prediction

S. Joe Qin^*

^*Corresponding author for this work

Department of Data Science

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

7 Citations (Scopus)

Abstract

In this paper, we propose a novel latent vector autoregressive (LaVAR) modeling algorithm with a canonical correlation analysis (CCA) objective to estimate a fully-interacting reduced dimensional dynamic model. This algorithm is an advancement of the dynamic inner canonical correlation analysis (DiCCA) algorithm, which builds univariate latent autoregressive models that are non-interacting. The dynamic latent variable scores of the proposed algorithm are enforced to be orthogonal or contemporaneously independent, similar to those of DiCCA. An application case study on an industrial dataset is given to illustrate the superiority of the proposed algorithm. The reduced-dimensional latent dynamic model has potential applications for prediction, control, and diagnosis of systems with rich sensors, such as industrial internet of things.

Original language	English
Title of host publication	2021 60th IEEE Conference on Decision and Control (CDC)
Publisher	IEEE
Pages	3689-3694
ISBN (Electronic)	9781665436595, 978-1-6654-3658-8
ISBN (Print)	978-1-6654-3660-1
DOIs	https://doi.org/10.1109/CDC45484.2021.9683255
Publication status	Published - Dec 2021
Event	60th IEEE Conference on Decision and Control (CDC 2021) - Fairmont Hotel (Virtual), Austin, United States Duration: 13 Dec 2021 → 17 Dec 2021 https://2021.ieeecdc.org/

Publication series

Name	Proceedings of the IEEE Conference on Decision and Control
ISSN (Print)	0743-1546
ISSN (Electronic)	2576-2370

Conference

Conference	60th IEEE Conference on Decision and Control (CDC 2021)
Place	United States
City	Austin
Period	13/12/21 → 17/12/21
Internet address	https://2021.ieeecdc.org/

UN SDGs

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

SDG 9 Industry, Innovation, and Infrastructure

Access Output

10.1109/CDC45484.2021.9683255

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

@inproceedings{f3ffdd4293634c749d063672e6ee9d56,

title = "Latent Vector Autoregressive Modeling for Reduced Dimensional Dynamic Feature Extraction and Prediction",

abstract = "In this paper, we propose a novel latent vector autoregressive (LaVAR) modeling algorithm with a canonical correlation analysis (CCA) objective to estimate a fully-interacting reduced dimensional dynamic model. This algorithm is an advancement of the dynamic inner canonical correlation analysis (DiCCA) algorithm, which builds univariate latent autoregressive models that are non-interacting. The dynamic latent variable scores of the proposed algorithm are enforced to be orthogonal or contemporaneously independent, similar to those of DiCCA. An application case study on an industrial dataset is given to illustrate the superiority of the proposed algorithm. The reduced-dimensional latent dynamic model has potential applications for prediction, control, and diagnosis of systems with rich sensors, such as industrial internet of things.",

author = "Qin, \{S. Joe\}",

year = "2021",

month = dec,

doi = "10.1109/CDC45484.2021.9683255",

language = "English",

isbn = "978-1-6654-3660-1",

series = "Proceedings of the IEEE Conference on Decision and Control",

publisher = "IEEE",

pages = "3689--3694",

booktitle = "2021 60th IEEE Conference on Decision and Control (CDC)",

address = "United States",

note = "60th IEEE Conference on Decision and Control (CDC 2021) ; Conference date: 13-12-2021 Through 17-12-2021",

url = "https://2021.ieeecdc.org/",

}

Qin, SJ 2021, Latent Vector Autoregressive Modeling for Reduced Dimensional Dynamic Feature Extraction and Prediction. in 2021 60th IEEE Conference on Decision and Control (CDC). Proceedings of the IEEE Conference on Decision and Control, IEEE, pp. 3689-3694, 60th IEEE Conference on Decision and Control (CDC 2021), Austin, United States, 13/12/21. https://doi.org/10.1109/CDC45484.2021.9683255

Latent Vector Autoregressive Modeling for Reduced Dimensional Dynamic Feature Extraction and Prediction. / Qin, S. Joe.
2021 60th IEEE Conference on Decision and Control (CDC). IEEE, 2021. p. 3689-3694 (Proceedings of the IEEE Conference on Decision and Control).

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

TY - GEN

T1 - Latent Vector Autoregressive Modeling for Reduced Dimensional Dynamic Feature Extraction and Prediction

AU - Qin, S. Joe

PY - 2021/12

Y1 - 2021/12

N2 - In this paper, we propose a novel latent vector autoregressive (LaVAR) modeling algorithm with a canonical correlation analysis (CCA) objective to estimate a fully-interacting reduced dimensional dynamic model. This algorithm is an advancement of the dynamic inner canonical correlation analysis (DiCCA) algorithm, which builds univariate latent autoregressive models that are non-interacting. The dynamic latent variable scores of the proposed algorithm are enforced to be orthogonal or contemporaneously independent, similar to those of DiCCA. An application case study on an industrial dataset is given to illustrate the superiority of the proposed algorithm. The reduced-dimensional latent dynamic model has potential applications for prediction, control, and diagnosis of systems with rich sensors, such as industrial internet of things.

AB - In this paper, we propose a novel latent vector autoregressive (LaVAR) modeling algorithm with a canonical correlation analysis (CCA) objective to estimate a fully-interacting reduced dimensional dynamic model. This algorithm is an advancement of the dynamic inner canonical correlation analysis (DiCCA) algorithm, which builds univariate latent autoregressive models that are non-interacting. The dynamic latent variable scores of the proposed algorithm are enforced to be orthogonal or contemporaneously independent, similar to those of DiCCA. An application case study on an industrial dataset is given to illustrate the superiority of the proposed algorithm. The reduced-dimensional latent dynamic model has potential applications for prediction, control, and diagnosis of systems with rich sensors, such as industrial internet of things.

UR - https://www.scopus.com/pages/publications/85126053592

UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-85126053592&origin=recordpage

U2 - 10.1109/CDC45484.2021.9683255

DO - 10.1109/CDC45484.2021.9683255

M3 - RGC 32 - Refereed conference paper (with host publication)

SN - 978-1-6654-3660-1

T3 - Proceedings of the IEEE Conference on Decision and Control

SP - 3689

EP - 3694

BT - 2021 60th IEEE Conference on Decision and Control (CDC)

PB - IEEE

T2 - 60th IEEE Conference on Decision and Control (CDC 2021)

Y2 - 13 December 2021 through 17 December 2021

ER -

Latent Vector Autoregressive Modeling for Reduced Dimensional Dynamic Feature Extraction and Prediction

Abstract

Publication series

Conference

UN SDGs

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Cite this