Classic Kriging versus Kriging with Bootstrapping or Conditional Simulation

Classic Kriging's Robust Confidence Intervals and Optimization (Revised version of CentER DP 2013-038)

Research output: Working paperDiscussion paperOther research output

650 Downloads (Pure)

Abstract

Kriging is a popular method for estimating the global optimum of a simulated system. Kriging approximates the input/output function of the simulation model. Kriging also estimates the variances of the predictions of outputs for input combinations not yet simulated. These predictions and their variances are used by efficient global optimization"(EGO), to balance local and global search. This article focuses on two related questions: (1) How to select the next combination to be simulated when searching for the global optimum? (2) How to derive confidence intervals for outputs of input combinations not yet simulated? Classic Kriging simply plugs the estimated Kriging parameters into the formula for the predictor variance, so theoretically this variance is biased. This article concludes that practitioners may ignore this bias, because classic Kriging gives acceptable confidence intervals and estimates of the optimal input
combination. This conclusion is based on bootstrapping and conditional
simulation.
Original languageEnglish
Place of PublicationTilburg
PublisherInformation Management
Number of pages22
Volume2014-076
Publication statusPublished - 2 Dec 2014

Publication series

NameCentER Discussion Paper
Volume2014-076

Fingerprint

bootstrapping
kriging
confidence interval
simulation
prediction

Keywords

  • simulaiton
  • Optimization
  • Kriging
  • Bootstrap
  • Conditional simulation

Cite this

@techreport{4915047bafe44fc78a1c45b8537fb4ea,
title = "Classic Kriging versus Kriging with Bootstrapping or Conditional Simulation: Classic Kriging's Robust Confidence Intervals and Optimization (Revised version of CentER DP 2013-038)",
abstract = "Kriging is a popular method for estimating the global optimum of a simulated system. Kriging approximates the input/output function of the simulation model. Kriging also estimates the variances of the predictions of outputs for input combinations not yet simulated. These predictions and their variances are used by efficient global optimization{"}(EGO), to balance local and global search. This article focuses on two related questions: (1) How to select the next combination to be simulated when searching for the global optimum? (2) How to derive confidence intervals for outputs of input combinations not yet simulated? Classic Kriging simply plugs the estimated Kriging parameters into the formula for the predictor variance, so theoretically this variance is biased. This article concludes that practitioners may ignore this bias, because classic Kriging gives acceptable confidence intervals and estimates of the optimal inputcombination. This conclusion is based on bootstrapping and conditionalsimulation.",
keywords = "simulaiton, Optimization, Kriging, Bootstrap, Conditional simulation",
author = "E. Mehdad and Kleijnen, {Jack P.C.}",
year = "2014",
month = "12",
day = "2",
language = "English",
volume = "2014-076",
series = "CentER Discussion Paper",
publisher = "Information Management",
type = "WorkingPaper",
institution = "Information Management",

}

Classic Kriging versus Kriging with Bootstrapping or Conditional Simulation : Classic Kriging's Robust Confidence Intervals and Optimization (Revised version of CentER DP 2013-038). / Mehdad, E.; Kleijnen, Jack P.C.

Tilburg : Information Management, 2014. (CentER Discussion Paper; Vol. 2014-076).

Research output: Working paperDiscussion paperOther research output

TY - UNPB

T1 - Classic Kriging versus Kriging with Bootstrapping or Conditional Simulation

T2 - Classic Kriging's Robust Confidence Intervals and Optimization (Revised version of CentER DP 2013-038)

AU - Mehdad, E.

AU - Kleijnen, Jack P.C.

PY - 2014/12/2

Y1 - 2014/12/2

N2 - Kriging is a popular method for estimating the global optimum of a simulated system. Kriging approximates the input/output function of the simulation model. Kriging also estimates the variances of the predictions of outputs for input combinations not yet simulated. These predictions and their variances are used by efficient global optimization"(EGO), to balance local and global search. This article focuses on two related questions: (1) How to select the next combination to be simulated when searching for the global optimum? (2) How to derive confidence intervals for outputs of input combinations not yet simulated? Classic Kriging simply plugs the estimated Kriging parameters into the formula for the predictor variance, so theoretically this variance is biased. This article concludes that practitioners may ignore this bias, because classic Kriging gives acceptable confidence intervals and estimates of the optimal inputcombination. This conclusion is based on bootstrapping and conditionalsimulation.

AB - Kriging is a popular method for estimating the global optimum of a simulated system. Kriging approximates the input/output function of the simulation model. Kriging also estimates the variances of the predictions of outputs for input combinations not yet simulated. These predictions and their variances are used by efficient global optimization"(EGO), to balance local and global search. This article focuses on two related questions: (1) How to select the next combination to be simulated when searching for the global optimum? (2) How to derive confidence intervals for outputs of input combinations not yet simulated? Classic Kriging simply plugs the estimated Kriging parameters into the formula for the predictor variance, so theoretically this variance is biased. This article concludes that practitioners may ignore this bias, because classic Kriging gives acceptable confidence intervals and estimates of the optimal inputcombination. This conclusion is based on bootstrapping and conditionalsimulation.

KW - simulaiton

KW - Optimization

KW - Kriging

KW - Bootstrap

KW - Conditional simulation

M3 - Discussion paper

VL - 2014-076

T3 - CentER Discussion Paper

BT - Classic Kriging versus Kriging with Bootstrapping or Conditional Simulation

PB - Information Management

CY - Tilburg

ER -