21734 articles – 15570 references  [version française]

hal-00578550, version 2

Optimization of Noisy Computer Experiments with Tunable Precision

Victor Picheny () 1, David Ginsbourger () 2, Yann Richet 3, Grégory Caplin 3

Abstract: This article addresses the issue of kriging-based optimization of stochastic simulators. Many of these simulators depend on factors that tune the level of precision of the response, the gain in accuracy being at a price of computational time. The contribution of this work is two-fold: firstly, we propose a quantile-based criterion for the sequential choice of experiments, in the fashion of the classical Expected Improvement criterion, which allows an elegant treatment of heterogeneous response precisions. Secondly, we present a procedure that allocates on-line the computational time given to each measurement, allowing a better distribution of the computational effort and increased efficiency. Finally, the optimization method is applied to an original application in nuclear criticality safety.

  • 1:  Mathématiques Appliquées aux Systèmes - EA 4037 (MAS)
  • Ecole Centrale Paris
  • 2:  Institute of Mathematical Statistics and Actuarial Science [Bern] (IMSV)
  • University of Bern
  • 3:  Institut de radioprotection et de sûreté nucléaire (IRSN)
  • Ministère de l'écologie de l'Energie, du Développement durable et de l'Aménagement du territoire – Ministère de l'économie, de l'industrie et de l'emploi – Ministère de l'Enseignement Supérieur et de la Recherche Scientifique – Ministère de la Défense – Ministère de la santé
 
  • hal-00578550, version 2
  • http://hal.archives-ouvertes.fr/hal-00578550
  • oai:hal.archives-ouvertes.fr:hal-00578550
  • From: Victor Picheny <>
  • Submitted on: Wednesday, 17 August 2011 09:53:12
  • Updated on: Wednesday, 17 August 2011 10:44:23