Using Gaussian process to enhance support vector regression

Yi Zhang; Wen Yao; Xiaoqian Chen; Fred van Keulen

doi:10.1007/978-3-319-67988-4_20

Using Gaussian process to enhance support vector regression

Yi Zhang, Wen Yao, Xiaoqian Chen, Fred van Keulen

Computational Design and Mechanics

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

5 Citations (Scopus)

Abstract

Support vector regression (SVR) is a common surrogate model for computationally expensive simulation. It is able to balance the model complexity and the error tolerance. Whether SVR interpolates the training samples is dependent on its parameters. For the nonlinear function approximation without noise, when SVR is not an interpolator, it is advisable to model the errors and use them to compensate the prediction response. In this paper, the errors of SVR are modeled by using Gaussian process, and the final model response is obtained by the combination of SVR and the Gaussian process of the errors. The numerical experiments show the proposed method is able to further improve the prediction accuracy of SVR.

Original language	English
Title of host publication	Advances in Structural and Multidisciplinary Optimization
Subtitle of host publication	Proceedings of the 12th World Congress of Structural and Multidisciplinary Optimization (WCSMO12)
Editors	A Schumacher, T Vietor, S Fiebig, K-U Bletzinger, K Maute
Place of Publication	Cham, Switzerland
Publisher	Springer
Pages	281-286
ISBN (Electronic)	978-3-319-67988-4
ISBN (Print)	978-3-319-67987-7
DOIs	https://doi.org/10.1007/978-3-319-67988-4_20
Publication status	Published - 2017
Event	WCSMO 2017: 12th World Congress of Structural and Multidisciplinary Optimisation - Braunschweig, Germany Duration: 5 Jun 2017 → 9 Jun 2017

Conference

Conference	WCSMO 2017: 12th World Congress of Structural and Multidisciplinary Optimisation
Country/Territory	Germany
City	Braunschweig
Period	5/06/17 → 9/06/17

Keywords

Support vector regression
Gaussian kernel
Error modeling
Gaussian process

Access to Document

10.1007/978-3-319-67988-4_20

https://link.springer.com/chapter/10.1007/978-3-319-67988-4_20#aboutcontent

Cite this

Zhang, Y., Yao, W., Chen, X., & van Keulen, F. (2017). Using Gaussian process to enhance support vector regression. In A. Schumacher, T. Vietor, S. Fiebig, K.-U. Bletzinger, & K. Maute (Eds.), Advances in Structural and Multidisciplinary Optimization : Proceedings of the 12th World Congress of Structural and Multidisciplinary Optimization (WCSMO12) (pp. 281-286). Springer. https://doi.org/10.1007/978-3-319-67988-4_20

Zhang, Yi ; Yao, Wen ; Chen, Xiaoqian et al. / Using Gaussian process to enhance support vector regression. Advances in Structural and Multidisciplinary Optimization : Proceedings of the 12th World Congress of Structural and Multidisciplinary Optimization (WCSMO12). editor / A Schumacher ; T Vietor ; S Fiebig ; K-U Bletzinger ; K Maute. Cham, Switzerland : Springer, 2017. pp. 281-286

@inproceedings{c7998f697f8d470abad13c65f6a862c9,

title = "Using Gaussian process to enhance support vector regression",

abstract = "Support vector regression (SVR) is a common surrogate model for computationally expensive simulation. It is able to balance the model complexity and the error tolerance. Whether SVR interpolates the training samples is dependent on its parameters. For the nonlinear function approximation without noise, when SVR is not an interpolator, it is advisable to model the errors and use them to compensate the prediction response. In this paper, the errors of SVR are modeled by using Gaussian process, and the final model response is obtained by the combination of SVR and the Gaussian process of the errors. The numerical experiments show the proposed method is able to further improve the prediction accuracy of SVR.",

keywords = "Support vector regression , Gaussian kernel, Error modeling, Gaussian process",

author = "Yi Zhang and Wen Yao and Xiaoqian Chen and {van Keulen}, Fred",

year = "2017",

doi = "10.1007/978-3-319-67988-4_20",

language = "English",

isbn = "978-3-319-67987-7 ",

pages = "281--286",

editor = "A Schumacher and T Vietor and S Fiebig and Bletzinger, {K-U } and K Maute",

booktitle = "Advances in Structural and Multidisciplinary Optimization",

publisher = "Springer",

note = "WCSMO 2017: 12th World Congress of Structural and Multidisciplinary Optimisation ; Conference date: 05-06-2017 Through 09-06-2017",

}

Zhang, Y, Yao, W, Chen, X & van Keulen, F 2017, Using Gaussian process to enhance support vector regression. in A Schumacher, T Vietor, S Fiebig, K-U Bletzinger & K Maute (eds), Advances in Structural and Multidisciplinary Optimization : Proceedings of the 12th World Congress of Structural and Multidisciplinary Optimization (WCSMO12). Springer, Cham, Switzerland, pp. 281-286, WCSMO 2017: 12th World Congress of Structural and Multidisciplinary Optimisation , Braunschweig, Germany, 5/06/17. https://doi.org/10.1007/978-3-319-67988-4_20

Using Gaussian process to enhance support vector regression. / Zhang, Yi; Yao, Wen; Chen, Xiaoqian et al.
Advances in Structural and Multidisciplinary Optimization : Proceedings of the 12th World Congress of Structural and Multidisciplinary Optimization (WCSMO12). ed. / A Schumacher; T Vietor; S Fiebig; K-U Bletzinger; K Maute. Cham, Switzerland: Springer, 2017. p. 281-286.

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

TY - GEN

T1 - Using Gaussian process to enhance support vector regression

AU - Zhang, Yi

AU - Yao, Wen

AU - Chen, Xiaoqian

AU - van Keulen, Fred

PY - 2017

Y1 - 2017

N2 - Support vector regression (SVR) is a common surrogate model for computationally expensive simulation. It is able to balance the model complexity and the error tolerance. Whether SVR interpolates the training samples is dependent on its parameters. For the nonlinear function approximation without noise, when SVR is not an interpolator, it is advisable to model the errors and use them to compensate the prediction response. In this paper, the errors of SVR are modeled by using Gaussian process, and the final model response is obtained by the combination of SVR and the Gaussian process of the errors. The numerical experiments show the proposed method is able to further improve the prediction accuracy of SVR.

AB - Support vector regression (SVR) is a common surrogate model for computationally expensive simulation. It is able to balance the model complexity and the error tolerance. Whether SVR interpolates the training samples is dependent on its parameters. For the nonlinear function approximation without noise, when SVR is not an interpolator, it is advisable to model the errors and use them to compensate the prediction response. In this paper, the errors of SVR are modeled by using Gaussian process, and the final model response is obtained by the combination of SVR and the Gaussian process of the errors. The numerical experiments show the proposed method is able to further improve the prediction accuracy of SVR.

KW - Support vector regression

KW - Gaussian kernel

KW - Error modeling

KW - Gaussian process

U2 - 10.1007/978-3-319-67988-4_20

DO - 10.1007/978-3-319-67988-4_20

M3 - Conference contribution

SN - 978-3-319-67987-7

SP - 281

EP - 286

BT - Advances in Structural and Multidisciplinary Optimization

A2 - Schumacher, A

A2 - Vietor, T

A2 - Fiebig, S

A2 - Bletzinger, K-U

A2 - Maute, K

PB - Springer

CY - Cham, Switzerland

T2 - WCSMO 2017: 12th World Congress of Structural and Multidisciplinary Optimisation

Y2 - 5 June 2017 through 9 June 2017

ER -

Zhang Y, Yao W, Chen X, van Keulen F. Using Gaussian process to enhance support vector regression. In Schumacher A, Vietor T, Fiebig S, Bletzinger KU, Maute K, editors, Advances in Structural and Multidisciplinary Optimization : Proceedings of the 12th World Congress of Structural and Multidisciplinary Optimization (WCSMO12). Cham, Switzerland: Springer. 2017. p. 281-286 doi: 10.1007/978-3-319-67988-4_20

Using Gaussian process to enhance support vector regression

Abstract

Conference

Keywords

Access to Document

Fingerprint

Cite this