Efficient Global Optimization for Black-Box Simulation via Sequential Intrinsic Kriging

Ehsan Mehdad, J.P.C. Kleijnen

Research output: Working paperDiscussion paperOther research output

572 Downloads (Pure)

Abstract

Efficient Global Optimization (EGO) is a popular method that searches sequentially for the global optimum of a simulated system. EGO treats the simulation model as a black-box, and balances local and global searches. In deterministic simulation, EGO uses ordinary Kriging (OK), which is a special case of universal Kriging (UK). In our EGO variant we use intrinsic Kriging (IK), which eliminates the need to estimate the parameters that quantify the trend in UK. In random simulation, EGO uses stochastic Kriging (SK), but we use stochastic IK (SIK). Moreover, in random simulation, EGO needs to select the number of replications per simulated input combination, accounting for the heteroscedastic variances of the simulation outputs. A popular selection method uses optimal computer budget allocation (OCBA), which allocates the available total number of replications over simulated combinations. We derive a new allocation algorithm. We perform several numerical experiments with deterministic simulations and random simulations. These experiments suggest that (1) in deterministic simulations, EGO with IK outperforms classic EGO; (2) in random simulations, EGO with SIK and our allocation rule does not differ significantly from EGO with SK combined with the OCBA allocation rule.
Original languageEnglish
Place of PublicationTilburg
PublisherDepartment of Econometrics and Operations Research
Number of pages18
Volume2015-042
Publication statusPublished - 18 Aug 2015

Publication series

NameCentER Discussion Paper
Volume2015-042

Fingerprint

kriging
simulation
experiment
allocation

Keywords

  • global optimization
  • Gaussian process
  • Kriging
  • intrinsic Krgigin
  • metamodel

Cite this

Mehdad, E., & Kleijnen, J. P. C. (2015). Efficient Global Optimization for Black-Box Simulation via Sequential Intrinsic Kriging. (CentER Discussion Paper; Vol. 2015-042). Tilburg: Department of Econometrics and Operations Research.
Mehdad, Ehsan ; Kleijnen, J.P.C. / Efficient Global Optimization for Black-Box Simulation via Sequential Intrinsic Kriging. Tilburg : Department of Econometrics and Operations Research, 2015. (CentER Discussion Paper).
@techreport{5e785713146c4e5bb671f7eb4a8b7a41,
title = "Efficient Global Optimization for Black-Box Simulation via Sequential Intrinsic Kriging",
abstract = "Efficient Global Optimization (EGO) is a popular method that searches sequentially for the global optimum of a simulated system. EGO treats the simulation model as a black-box, and balances local and global searches. In deterministic simulation, EGO uses ordinary Kriging (OK), which is a special case of universal Kriging (UK). In our EGO variant we use intrinsic Kriging (IK), which eliminates the need to estimate the parameters that quantify the trend in UK. In random simulation, EGO uses stochastic Kriging (SK), but we use stochastic IK (SIK). Moreover, in random simulation, EGO needs to select the number of replications per simulated input combination, accounting for the heteroscedastic variances of the simulation outputs. A popular selection method uses optimal computer budget allocation (OCBA), which allocates the available total number of replications over simulated combinations. We derive a new allocation algorithm. We perform several numerical experiments with deterministic simulations and random simulations. These experiments suggest that (1) in deterministic simulations, EGO with IK outperforms classic EGO; (2) in random simulations, EGO with SIK and our allocation rule does not differ significantly from EGO with SK combined with the OCBA allocation rule.",
keywords = "global optimization, Gaussian process, Kriging, intrinsic Krgigin, metamodel",
author = "Ehsan Mehdad and J.P.C. Kleijnen",
year = "2015",
month = "8",
day = "18",
language = "English",
volume = "2015-042",
series = "CentER Discussion Paper",
publisher = "Department of Econometrics and Operations Research",
type = "WorkingPaper",
institution = "Department of Econometrics and Operations Research",

}

Mehdad, E & Kleijnen, JPC 2015 'Efficient Global Optimization for Black-Box Simulation via Sequential Intrinsic Kriging' CentER Discussion Paper, vol. 2015-042, Department of Econometrics and Operations Research, Tilburg.

Efficient Global Optimization for Black-Box Simulation via Sequential Intrinsic Kriging. / Mehdad, Ehsan; Kleijnen, J.P.C.

Tilburg : Department of Econometrics and Operations Research, 2015. (CentER Discussion Paper; Vol. 2015-042).

Research output: Working paperDiscussion paperOther research output

TY - UNPB

T1 - Efficient Global Optimization for Black-Box Simulation via Sequential Intrinsic Kriging

AU - Mehdad, Ehsan

AU - Kleijnen, J.P.C.

PY - 2015/8/18

Y1 - 2015/8/18

N2 - Efficient Global Optimization (EGO) is a popular method that searches sequentially for the global optimum of a simulated system. EGO treats the simulation model as a black-box, and balances local and global searches. In deterministic simulation, EGO uses ordinary Kriging (OK), which is a special case of universal Kriging (UK). In our EGO variant we use intrinsic Kriging (IK), which eliminates the need to estimate the parameters that quantify the trend in UK. In random simulation, EGO uses stochastic Kriging (SK), but we use stochastic IK (SIK). Moreover, in random simulation, EGO needs to select the number of replications per simulated input combination, accounting for the heteroscedastic variances of the simulation outputs. A popular selection method uses optimal computer budget allocation (OCBA), which allocates the available total number of replications over simulated combinations. We derive a new allocation algorithm. We perform several numerical experiments with deterministic simulations and random simulations. These experiments suggest that (1) in deterministic simulations, EGO with IK outperforms classic EGO; (2) in random simulations, EGO with SIK and our allocation rule does not differ significantly from EGO with SK combined with the OCBA allocation rule.

AB - Efficient Global Optimization (EGO) is a popular method that searches sequentially for the global optimum of a simulated system. EGO treats the simulation model as a black-box, and balances local and global searches. In deterministic simulation, EGO uses ordinary Kriging (OK), which is a special case of universal Kriging (UK). In our EGO variant we use intrinsic Kriging (IK), which eliminates the need to estimate the parameters that quantify the trend in UK. In random simulation, EGO uses stochastic Kriging (SK), but we use stochastic IK (SIK). Moreover, in random simulation, EGO needs to select the number of replications per simulated input combination, accounting for the heteroscedastic variances of the simulation outputs. A popular selection method uses optimal computer budget allocation (OCBA), which allocates the available total number of replications over simulated combinations. We derive a new allocation algorithm. We perform several numerical experiments with deterministic simulations and random simulations. These experiments suggest that (1) in deterministic simulations, EGO with IK outperforms classic EGO; (2) in random simulations, EGO with SIK and our allocation rule does not differ significantly from EGO with SK combined with the OCBA allocation rule.

KW - global optimization

KW - Gaussian process

KW - Kriging

KW - intrinsic Krgigin

KW - metamodel

M3 - Discussion paper

VL - 2015-042

T3 - CentER Discussion Paper

BT - Efficient Global Optimization for Black-Box Simulation via Sequential Intrinsic Kriging

PB - Department of Econometrics and Operations Research

CY - Tilburg

ER -

Mehdad E, Kleijnen JPC. Efficient Global Optimization for Black-Box Simulation via Sequential Intrinsic Kriging. Tilburg: Department of Econometrics and Operations Research. 2015 Aug 18. (CentER Discussion Paper).