Kriging-based optimization algorithms for noisy data

Resumo

Responses to many real-world problems can only be evaluated perturbed by noise. Intelligent opti-
mization strategies, successfully coping with noisy evaluations, are required in order to enable making efficient

optimization of these problems. The surrogate model has been popularly used in the area of design optimization
with high computational cost, especially in Kriging-based optimization algorithms. The performance of those
algorithms depends on a sequential search of so-called infill points, used to update the Kriging meta-model at

each iteration. This article explores the most relevant single and multi-objective infill algorithms used for Kriging-
based optimization with noise-handling strategies. Those algorithms explore information about the variance of the

predictor and the noise from stochastic simulation.