### Abstract

Original language | English |
---|---|

Title of host publication | Proceedings of the 2012 Winter Simulation Conference |

Editors | C. Laroque, J. Himmelspach, R. Pasupathy, O. Rose, A.M. Uhrmacher |

Place of Publication | Berlin |

Publisher | Unknown Publisher |

Pages | 543-554 |

Publication status | Published - 2012 |

### Fingerprint

### Cite this

*Proceedings of the 2012 Winter Simulation Conference*(pp. 543-554). Berlin: Unknown Publisher.

}

*Proceedings of the 2012 Winter Simulation Conference.*Unknown Publisher, Berlin, pp. 543-554.

**Convex and monotonic bootstrapped kriging.** / Kleijnen, Jack P.C.; Mehdad, E.; van Beers, W.C.M.

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

TY - GEN

T1 - Convex and monotonic bootstrapped kriging

AU - Kleijnen, Jack P.C.

AU - Mehdad, E.

AU - van Beers, W.C.M.

PY - 2012

Y1 - 2012

N2 - Distribution-free bootstrapping of the replicated responses of a given discrete-event simulation model gives bootstrapped Kriging (Gaussian process) metamodels; we require these metamodels to be either convex or monotonic. To illustrate monotonic Kriging, we use an M/M/1 queueing simulation with as output either the mean or the 90% quantile of the transient-state waiting times, and as input the traffic rate. In this example, monotonic bootstrapped Kriging enables better sensitivity analysis than classic Kriging; i.e., bootstrapping gives lower MSE and confidence intervals with higher coverage and the same length. To illustrate convex Kriging, we start with simulation-optimization of an (s, S) inventory model, but we next switch to a Monte Carlo experiment with a second-order polynomial inspired by this inventory simulation. We could not find truly convex Kriging metamodels, either classic or bootstrapped; nevertheless, our bootstrapped “nearly convex” Kriging does give a confidence interval for the optimal input combination.

AB - Distribution-free bootstrapping of the replicated responses of a given discrete-event simulation model gives bootstrapped Kriging (Gaussian process) metamodels; we require these metamodels to be either convex or monotonic. To illustrate monotonic Kriging, we use an M/M/1 queueing simulation with as output either the mean or the 90% quantile of the transient-state waiting times, and as input the traffic rate. In this example, monotonic bootstrapped Kriging enables better sensitivity analysis than classic Kriging; i.e., bootstrapping gives lower MSE and confidence intervals with higher coverage and the same length. To illustrate convex Kriging, we start with simulation-optimization of an (s, S) inventory model, but we next switch to a Monte Carlo experiment with a second-order polynomial inspired by this inventory simulation. We could not find truly convex Kriging metamodels, either classic or bootstrapped; nevertheless, our bootstrapped “nearly convex” Kriging does give a confidence interval for the optimal input combination.

M3 - Conference contribution

SP - 543

EP - 554

BT - Proceedings of the 2012 Winter Simulation Conference

A2 - Laroque, C.

A2 - Himmelspach, J.

A2 - Pasupathy, R.

A2 - Rose, O.

A2 - Uhrmacher, A.M.

PB - Unknown Publisher

CY - Berlin

ER -