Simulation Optmization

Stochastic Kriging with Gradient

This is an implementation of the Stochastic Kriging with Gradient (SKG) proposed by Chen et al. (2013).

The codes were written in MATLAB R2015a. They (except for the third-party functions mentioned below) can be freely redistributed and used under the terms of the BSD 3-Clause License.

Download the entire package with an example in it.

Remarks

Gaussian (or called squared exponential) correlation function of the form $R(\boldsymbol{x}-\boldsymbol{y};\boldsymbol{\theta })=\textup{exp}(-\sum_{i=1}^{d}\theta_i |x_i-y_i|^2)$ with $\theta_i>0$ for $i=1,\ldots,d$ is used. It is easy to modify the function C in “neglogLL.m” and “SKGpredict.m” to use other types of correlation function.
By default, gradients along all coordinates are incorporated. Modification will be required to contain only part of the gradients.

» Go back to the homepage