New paper published in SIAM/ASA Journal of Uncertainty Quantification
X. Zhu and B. Sudret published a new paper on building surrogate models to represent the response distribution of stochastic simulators.
The paper entitled Emulation of stochastic simulators using generalized lambda models proposes a novel surrogate model called generalized lambda model to emulate stochastic simulators. This model uses the flexible generalized lambda distribution to approximate the response distribution of a stochastic simulator. The distribution parameters as functions of the input variables are represented by polynomial chaos expansions. The paper presents an adaptive method to build such a surrogate model without the need for replications.
For more information, please visit external pagethis for the publication and this for the associated report on our internal archive.