A Novel Two-step Sparse Learning Approach for Variable Selection and Optimal Predictive Modeling

Yiren Liu; S. Joe Qin

doi:10.1016/j.ifacol.2022.07.422

A Novel Two-step Sparse Learning Approach for Variable Selection and Optimal Predictive Modeling

Yiren Liu, S. Joe Qin^*

^*Corresponding author for this work

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

5 Citations (Scopus)

85 Downloads (CityUHK Scholars)

Abstract

In this paper, a two-step sparse learning approach is proposed for variable selection and model parameter estimation with optimally tuned hyperparameters in each step. In Step one, a sparse learning algorithm is applied on all data to produce a sequence of candidate subsets of selected variables by varying the hyperparameter value. In Step two, for each subset of the selected variables from Step one, Lasso, ridge regression, elastic-net, or adaptive Lasso is employed to find the optimal hyperparameters with the best cross-validation error. Among all subsets, the one with the overall minimum cross-validation error is selected as globally optimal. The effectiveness of the proposed approach is demonstrated using an industrial NOx emission dataset and the Dow challenge dataset to predict product impurity. Copyright (C) 2022 The Authors.

Original language	English
Title of host publication	IFAC-PapersOnLine
Publisher	International Federation of Automatic Control (IFAC)
Pages	57-64
Volume	55
DOIs	https://doi.org/10.1016/j.ifacol.2022.07.422
Publication status	Published - 2022
Event	13th IFAC Symposium on Dynamics and Control of Process Systems, including Biosystems (DYCOPS) - Busan, Korea, Republic of Duration: 14 Jun 2022 → 17 Jun 2022

Conference

Conference	13th IFAC Symposium on Dynamics and Control of Process Systems, including Biosystems (DYCOPS)
Place	Korea, Republic of
City	Busan
Period	14/06/22 → 17/06/22

Research Keywords

Inferential modeling
sparse statistical learning
variable selection
regularization
industrial applications
INFERENTIAL SENSORS
REGRESSION

Publisher's Copyright Statement

This full text is made available under CC-BY-NC-ND 4.0. https://creativecommons.org/licenses/by-nc-nd/4.0/

Access Output

10.1016/j.ifacol.2022.07.422

122933799.pdfFinal Published version, 2.67 MBLicence: CC BY-NC-ND

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

@inproceedings{b0bf5924dccd45edb5d41b9ddc855fb3,

title = "A Novel Two-step Sparse Learning Approach for Variable Selection and Optimal Predictive Modeling",

abstract = "In this paper, a two-step sparse learning approach is proposed for variable selection and model parameter estimation with optimally tuned hyperparameters in each step. In Step one, a sparse learning algorithm is applied on all data to produce a sequence of candidate subsets of selected variables by varying the hyperparameter value. In Step two, for each subset of the selected variables from Step one, Lasso, ridge regression, elastic-net, or adaptive Lasso is employed to find the optimal hyperparameters with the best cross-validation error. Among all subsets, the one with the overall minimum cross-validation error is selected as globally optimal. The effectiveness of the proposed approach is demonstrated using an industrial NOx emission dataset and the Dow challenge dataset to predict product impurity. Copyright (C) 2022 The Authors.",

keywords = "Inferential modeling, sparse statistical learning, variable selection, regularization, industrial applications, INFERENTIAL SENSORS, REGRESSION",

author = "Yiren Liu and Qin, {S. Joe}",

year = "2022",

doi = "10.1016/j.ifacol.2022.07.422",

language = "English",

volume = "55",

pages = "57--64",

booktitle = "IFAC-PapersOnLine",

publisher = "International Federation of Automatic Control (IFAC)",

note = "13th IFAC Symposium on Dynamics and Control of Process Systems, including Biosystems (DYCOPS) ; Conference date: 14-06-2022 Through 17-06-2022",

}

Liu, Y & Qin, SJ 2022, A Novel Two-step Sparse Learning Approach for Variable Selection and Optimal Predictive Modeling. in IFAC-PapersOnLine. vol. 55, International Federation of Automatic Control (IFAC), pp. 57-64, 13th IFAC Symposium on Dynamics and Control of Process Systems, including Biosystems (DYCOPS), Busan, Korea, Republic of, 14/06/22. https://doi.org/10.1016/j.ifacol.2022.07.422

A Novel Two-step Sparse Learning Approach for Variable Selection and Optimal Predictive Modeling. / Liu, Yiren; Qin, S. Joe.
IFAC-PapersOnLine. Vol. 55 International Federation of Automatic Control (IFAC), 2022. p. 57-64.

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

TY - GEN

T1 - A Novel Two-step Sparse Learning Approach for Variable Selection and Optimal Predictive Modeling

AU - Liu, Yiren

AU - Qin, S. Joe

PY - 2022

Y1 - 2022

N2 - In this paper, a two-step sparse learning approach is proposed for variable selection and model parameter estimation with optimally tuned hyperparameters in each step. In Step one, a sparse learning algorithm is applied on all data to produce a sequence of candidate subsets of selected variables by varying the hyperparameter value. In Step two, for each subset of the selected variables from Step one, Lasso, ridge regression, elastic-net, or adaptive Lasso is employed to find the optimal hyperparameters with the best cross-validation error. Among all subsets, the one with the overall minimum cross-validation error is selected as globally optimal. The effectiveness of the proposed approach is demonstrated using an industrial NOx emission dataset and the Dow challenge dataset to predict product impurity. Copyright (C) 2022 The Authors.

AB - In this paper, a two-step sparse learning approach is proposed for variable selection and model parameter estimation with optimally tuned hyperparameters in each step. In Step one, a sparse learning algorithm is applied on all data to produce a sequence of candidate subsets of selected variables by varying the hyperparameter value. In Step two, for each subset of the selected variables from Step one, Lasso, ridge regression, elastic-net, or adaptive Lasso is employed to find the optimal hyperparameters with the best cross-validation error. Among all subsets, the one with the overall minimum cross-validation error is selected as globally optimal. The effectiveness of the proposed approach is demonstrated using an industrial NOx emission dataset and the Dow challenge dataset to predict product impurity. Copyright (C) 2022 The Authors.

KW - Inferential modeling

KW - sparse statistical learning

KW - variable selection

KW - regularization

KW - industrial applications

KW - INFERENTIAL SENSORS

KW - REGRESSION

UR - http://gateway.isiknowledge.com/gateway/Gateway.cgi?GWVersion=2&SrcAuth=LinksAMR&SrcApp=PARTNER_APP&DestLinkType=FullRecord&DestApp=WOS&KeyUT=000876128600010

UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-85137034794&origin=recordpage

UR - http://www.scopus.com/inward/record.url?scp=85137034794&partnerID=8YFLogxK

U2 - 10.1016/j.ifacol.2022.07.422

DO - 10.1016/j.ifacol.2022.07.422

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

VL - 55

SP - 57

EP - 64

BT - IFAC-PapersOnLine

PB - International Federation of Automatic Control (IFAC)

T2 - 13th IFAC Symposium on Dynamics and Control of Process Systems, including Biosystems (DYCOPS)

Y2 - 14 June 2022 through 17 June 2022

ER -

A Novel Two-step Sparse Learning Approach for Variable Selection and Optimal Predictive Modeling

Abstract

Conference

Research Keywords

Publisher's Copyright Statement

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Cite this