Stochastic kriging is a metamodeling methodology developed for stochastic simulation experiments; it is based on the highly successful kriging method for the design and analysis of computer experiments. Stochastic kriging distinguishes the uncertainty about the performance response surface from the sampling uncertainty inherent in the stochastic simulation; it accomplishes this by incorporating trend models that are common in least-squares regression, spatial correlation modeling to account for inadequacies of the trend model, and an output-variance model which is also based on spatial correlation concepts. Stochastic kriging facilitates adaptive, sequential experiment designs that systematically reduce both model and sampling uncertainty to a user-specified level.

The stochastic kriging Matlab software available from this web site is distributed "as is," without warranties of any kind, either express or implied. Altough no user support is provide, please report problems via the contact link.

The software is copyrighted by Barry L. Nelson, Jeremy Staum, Evren Baysal and Wei Xie 2009. The authors grant permission for unlimited personal use of this software without fee. However, no derivative works based on the software may be prepared, including embedding any portion of it in another software product, without permission of the copyright holders.

Last update 10/24/2010