Comparison of Kriging-based algorithms for optimization with heterogeneous noise

Resumo

Problem modeling through response surfaces, or meta-models, has been a great solution adopted for opti-
mizing problems with high computational cost, especially Kriging-based optimization algorithms. In recent years,

algorithms have been proposed which extend the traditional Kriging-based simulation optimization algorithms
(assuming deterministic outputs) to problems in the presence of noise or uncertainty. This paper approaching
stochastic kriging meta-model in a comparative study of the performance of three Kriging-based algorithms for
unconstrained minimization a noisy function. The Minimum Quantile criterion (MQ), stochastic Efficient Global
Optimization (sEGO) and Expected Improvement with Reinterpolation (EIR) will be the algorithms compared

using an analytical test function. The conclusions and insights obtained may serve as a useful guideline for re-
searchers aiming to deal with optimization problems, especially to apply Kriging-based algorithms to solve engi-
neering problems, and may be useful in the development of future algorithms.