A new combined sampling method based on variance minimization strategy

Xiaotian Xie; Wei Li; Lingyun Lu; Ming Yang

doi:10.1109/CCDC.2016.7531281

A new combined sampling method based on variance minimization strategy

Xiaotian Xie
, Wei Li
, Lingyun Lu
, Ming Yang

School of Astronautics

Harbin Institute of Technology

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

1 Scopus citations

Abstract

Due to the apparent drawbacks of Crude Monte Carlo method (CMC)-including requiring a large number of samples, comparatively lower sampling efficiency, an improved sampling technique based on Variance Minimization (VM) strategy is proposed. The VM strategy is a versatile and widely-applied adaptive importance sampling technique. In this paper, the VM strategy is applied to find an importance probability density function (IPDF) before sampling. And then the obtained IPDF is sampled by Latin Hypercube Sampling (LHS). Besides, the variance of test function can be further reduced through Antithetic Random Variables (ARV) technique. The simulation results of two examples show that this new combined sampling method based on VM strategy (CSMVMS) can effectively reduce sample size and enhance efficiency under certain level of precision.

Original language	English
Title of host publication	Proceedings of the 28th Chinese Control and Decision Conference, CCDC 2016
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	1841-1844
Number of pages	4
ISBN (Electronic)	9781467397148
DOIs	https://doi.org/10.1109/CCDC.2016.7531281
State	Published - 3 Aug 2016
Event	28th Chinese Control and Decision Conference, CCDC 2016 - Yinchuan, China Duration: 28 May 2016 → 30 May 2016

Publication series

Name	Proceedings of the 28th Chinese Control and Decision Conference, CCDC 2016

Conference

Conference	28th Chinese Control and Decision Conference, CCDC 2016
Country/Territory	China
City	Yinchuan
Period	28/05/16 → 30/05/16

Keywords

Antithetic Random Variables
Latin Hypercube Sampling
Variance Minimization strategy
combined sampling

Access to Document

10.1109/CCDC.2016.7531281

Cite this

Xie, X., Li, W., Lu, L., & Yang, M. (2016). A new combined sampling method based on variance minimization strategy. In Proceedings of the 28th Chinese Control and Decision Conference, CCDC 2016 (pp. 1841-1844). Article 7531281 (Proceedings of the 28th Chinese Control and Decision Conference, CCDC 2016). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/CCDC.2016.7531281

@inproceedings{cebe3296641044cda203cec2eb24300a,

title = "A new combined sampling method based on variance minimization strategy",

abstract = "Due to the apparent drawbacks of Crude Monte Carlo method (CMC)-including requiring a large number of samples, comparatively lower sampling efficiency, an improved sampling technique based on Variance Minimization (VM) strategy is proposed. The VM strategy is a versatile and widely-applied adaptive importance sampling technique. In this paper, the VM strategy is applied to find an importance probability density function (IPDF) before sampling. And then the obtained IPDF is sampled by Latin Hypercube Sampling (LHS). Besides, the variance of test function can be further reduced through Antithetic Random Variables (ARV) technique. The simulation results of two examples show that this new combined sampling method based on VM strategy (CSMVMS) can effectively reduce sample size and enhance efficiency under certain level of precision.",

keywords = "Antithetic Random Variables, Latin Hypercube Sampling, Variance Minimization strategy, combined sampling",

author = "Xiaotian Xie and Wei Li and Lingyun Lu and Ming Yang",

note = "Publisher Copyright: {\textcopyright} 2016 IEEE.; 28th Chinese Control and Decision Conference, CCDC 2016 ; Conference date: 28-05-2016 Through 30-05-2016",

year = "2016",

month = aug,

day = "3",

doi = "10.1109/CCDC.2016.7531281",

language = "英语",

series = "Proceedings of the 28th Chinese Control and Decision Conference, CCDC 2016",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "1841--1844",

booktitle = "Proceedings of the 28th Chinese Control and Decision Conference, CCDC 2016",

address = "美国",

}

Xie, X, Li, W, Lu, L & Yang, M 2016, A new combined sampling method based on variance minimization strategy. in Proceedings of the 28th Chinese Control and Decision Conference, CCDC 2016., 7531281, Proceedings of the 28th Chinese Control and Decision Conference, CCDC 2016, Institute of Electrical and Electronics Engineers Inc., pp. 1841-1844, 28th Chinese Control and Decision Conference, CCDC 2016, Yinchuan, China, 28/05/16. https://doi.org/10.1109/CCDC.2016.7531281

A new combined sampling method based on variance minimization strategy. / Xie, Xiaotian; Li, Wei; Lu, Lingyun et al.
Proceedings of the 28th Chinese Control and Decision Conference, CCDC 2016. Institute of Electrical and Electronics Engineers Inc., 2016. p. 1841-1844 7531281 (Proceedings of the 28th Chinese Control and Decision Conference, CCDC 2016).

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

TY - GEN

T1 - A new combined sampling method based on variance minimization strategy

AU - Xie, Xiaotian

AU - Li, Wei

AU - Lu, Lingyun

AU - Yang, Ming

PY - 2016/8/3

Y1 - 2016/8/3

N2 - Due to the apparent drawbacks of Crude Monte Carlo method (CMC)-including requiring a large number of samples, comparatively lower sampling efficiency, an improved sampling technique based on Variance Minimization (VM) strategy is proposed. The VM strategy is a versatile and widely-applied adaptive importance sampling technique. In this paper, the VM strategy is applied to find an importance probability density function (IPDF) before sampling. And then the obtained IPDF is sampled by Latin Hypercube Sampling (LHS). Besides, the variance of test function can be further reduced through Antithetic Random Variables (ARV) technique. The simulation results of two examples show that this new combined sampling method based on VM strategy (CSMVMS) can effectively reduce sample size and enhance efficiency under certain level of precision.

AB - Due to the apparent drawbacks of Crude Monte Carlo method (CMC)-including requiring a large number of samples, comparatively lower sampling efficiency, an improved sampling technique based on Variance Minimization (VM) strategy is proposed. The VM strategy is a versatile and widely-applied adaptive importance sampling technique. In this paper, the VM strategy is applied to find an importance probability density function (IPDF) before sampling. And then the obtained IPDF is sampled by Latin Hypercube Sampling (LHS). Besides, the variance of test function can be further reduced through Antithetic Random Variables (ARV) technique. The simulation results of two examples show that this new combined sampling method based on VM strategy (CSMVMS) can effectively reduce sample size and enhance efficiency under certain level of precision.

KW - Antithetic Random Variables

KW - Latin Hypercube Sampling

KW - Variance Minimization strategy

KW - combined sampling

UR - https://www.scopus.com/pages/publications/84983776361

U2 - 10.1109/CCDC.2016.7531281

DO - 10.1109/CCDC.2016.7531281

M3 - 会议稿件

AN - SCOPUS:84983776361

T3 - Proceedings of the 28th Chinese Control and Decision Conference, CCDC 2016

SP - 1841

EP - 1844

BT - Proceedings of the 28th Chinese Control and Decision Conference, CCDC 2016

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 28th Chinese Control and Decision Conference, CCDC 2016

Y2 - 28 May 2016 through 30 May 2016

ER -

Xie X, Li W, Lu L, Yang M. A new combined sampling method based on variance minimization strategy. In Proceedings of the 28th Chinese Control and Decision Conference, CCDC 2016. Institute of Electrical and Electronics Engineers Inc. 2016. p. 1841-1844. 7531281. (Proceedings of the 28th Chinese Control and Decision Conference, CCDC 2016). doi: 10.1109/CCDC.2016.7531281

A new combined sampling method based on variance minimization strategy

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this