Simulation Optimization through Regression or Kriging Metamodels

Research output: Working paperDiscussion paperOther research output

253 Downloads (Pure)

Abstract

This chapter surveys two methods for the optimization of real-world systems that are modelled through simulation. These methods use either linear regression metamodels, or Kriging (Gaussian processes). The metamodel type guides the design of the experiment; this design …fixes the input combinations of the simulation model. These regression models uses a sequence of local fi…rst-order and second-order polynomials— known as response surface methodology (RSM). Kriging models are global, but are re-estimated through sequential designs. "Robust" optimization may use RSM or Kriging, and accounts for uncertainty in simulation inputs.
Original languageEnglish
Place of PublicationTilburg
PublisherCentER, Center for Economic Research
Number of pages21
Volume2017-026
Publication statusPublished - 16 May 2017

Publication series

NameCentER Discussion Paper
Volume2017-026

Fingerprint

kriging
simulation
experiment
response surface methodology
method

Keywords

  • Cross-validation
  • Robust optimization
  • Regression analysis
  • Kriging
  • Guassian Process
  • response surface methodology (RSM)
  • efficient global optimization (EGO)
  • Taguchi
  • Boottrap
  • common rondom numbers (CRN)
  • Ltin hypercube sampling (LHS)
  • Karush-Kuhn-Tucker (KKT)

Cite this

Kleijnen, J. P. C. (2017). Simulation Optimization through Regression or Kriging Metamodels. (CentER Discussion Paper; Vol. 2017-026). Tilburg: CentER, Center for Economic Research.
Kleijnen, J.P.C. / Simulation Optimization through Regression or Kriging Metamodels. Tilburg : CentER, Center for Economic Research, 2017. (CentER Discussion Paper).
@techreport{c7f60f029dc541fc897f2e58c1824f01,
title = "Simulation Optimization through Regression or Kriging Metamodels",
abstract = "This chapter surveys two methods for the optimization of real-world systems that are modelled through simulation. These methods use either linear regression metamodels, or Kriging (Gaussian processes). The metamodel type guides the design of the experiment; this design …fixes the input combinations of the simulation model. These regression models uses a sequence of local fi…rst-order and second-order polynomials— known as response surface methodology (RSM). Kriging models are global, but are re-estimated through sequential designs. {"}Robust{"} optimization may use RSM or Kriging, and accounts for uncertainty in simulation inputs.",
keywords = "Cross-validation, Robust optimization, Regression analysis, Kriging, Guassian Process, response surface methodology (RSM), efficient global optimization (EGO), Taguchi, Boottrap, common rondom numbers (CRN), Ltin hypercube sampling (LHS), Karush-Kuhn-Tucker (KKT)",
author = "J.P.C. Kleijnen",
year = "2017",
month = "5",
day = "16",
language = "English",
volume = "2017-026",
series = "CentER Discussion Paper",
publisher = "CentER, Center for Economic Research",
type = "WorkingPaper",
institution = "CentER, Center for Economic Research",

}

Kleijnen, JPC 2017 'Simulation Optimization through Regression or Kriging Metamodels' CentER Discussion Paper, vol. 2017-026, CentER, Center for Economic Research, Tilburg.

Simulation Optimization through Regression or Kriging Metamodels. / Kleijnen, J.P.C.

Tilburg : CentER, Center for Economic Research, 2017. (CentER Discussion Paper; Vol. 2017-026).

Research output: Working paperDiscussion paperOther research output

TY - UNPB

T1 - Simulation Optimization through Regression or Kriging Metamodels

AU - Kleijnen, J.P.C.

PY - 2017/5/16

Y1 - 2017/5/16

N2 - This chapter surveys two methods for the optimization of real-world systems that are modelled through simulation. These methods use either linear regression metamodels, or Kriging (Gaussian processes). The metamodel type guides the design of the experiment; this design …fixes the input combinations of the simulation model. These regression models uses a sequence of local fi…rst-order and second-order polynomials— known as response surface methodology (RSM). Kriging models are global, but are re-estimated through sequential designs. "Robust" optimization may use RSM or Kriging, and accounts for uncertainty in simulation inputs.

AB - This chapter surveys two methods for the optimization of real-world systems that are modelled through simulation. These methods use either linear regression metamodels, or Kriging (Gaussian processes). The metamodel type guides the design of the experiment; this design …fixes the input combinations of the simulation model. These regression models uses a sequence of local fi…rst-order and second-order polynomials— known as response surface methodology (RSM). Kriging models are global, but are re-estimated through sequential designs. "Robust" optimization may use RSM or Kriging, and accounts for uncertainty in simulation inputs.

KW - Cross-validation

KW - Robust optimization

KW - Regression analysis

KW - Kriging

KW - Guassian Process

KW - response surface methodology (RSM)

KW - efficient global optimization (EGO)

KW - Taguchi

KW - Boottrap

KW - common rondom numbers (CRN)

KW - Ltin hypercube sampling (LHS)

KW - Karush-Kuhn-Tucker (KKT)

M3 - Discussion paper

VL - 2017-026

T3 - CentER Discussion Paper

BT - Simulation Optimization through Regression or Kriging Metamodels

PB - CentER, Center for Economic Research

CY - Tilburg

ER -

Kleijnen JPC. Simulation Optimization through Regression or Kriging Metamodels. Tilburg: CentER, Center for Economic Research. 2017 May 16. (CentER Discussion Paper).