A benchmark of kriging-based infill criteria for noisy optimization

Victor Picheny (1) , Tobias Wagner (2) , David Ginsbourger (3)

1 CERFACS - Centre Européen de Recherche et de Formation Avancée en Calcul Scientifique
2 TU - Technische Universität Dortmund [Dortmund]
3 IMSV - Institute of Mathematical Statistics and Actuarial Science [Bern]

Responses of many real-world problems can only be evaluated perturbed by noise. In order to make an efficient optimization of these problems possible, intelligent optimization strategies successfully coping with noisy evaluations are required. In this article, a comprehensive comparison of existing kriging-based methods for the optimization of noisy functions is provided. Ten methods are described using a unified formalism, and compared on analytical benchmark problems with different configurations (noise level, maximum number of observations, initial number of observations). It is found that the optimal method depends on the optimization problem, even though some criteria are consistently more efficient than others.

Mots clés: Gaussian process; EGO; noise

Domaines: Optimisation et contrôle [math.OC]

Victor Picheny, Tobias Wagner, David Ginsbourger. A benchmark of kriging-based infill criteria for noisy optimization. 2012. ⟨hal-00658212⟩