Gradient-enhanced kriging
Gradient-enhanced kriging (GEK) is a surrogate modeling technique used in engineering.
≈ 1 min read
Names in other languages
No names in other languages are held for this entry yet.
Article bodies are English and are not machine-translated. A name in another language is not a translation of the article.
Gradient-enhanced kriging (GEK) is a surrogate modeling technique used in engineering. A surrogate model (alternatively known as a metamodel, response surface or emulator) is a prediction of the output of an expensive computer code.
This prediction is based on a small number of evaluations of the expensive computer code. s = \frac{N+dN}{2N} = \frac{1}{2} + \frac{1}{2}d. A linear speedup has been demonstrated for a fluid-structure interaction problem and for a transonic airfoil. \operatorname E(X\mid y) = \mu + K(y-H\mu), and Kriging covariance: \operatorname{cov}(X\mid y) = (I-KH)P, where we have the gain matrix: K = PH^T(R+HPH^T)^{-1}. In Kriging, the prior covariance matrix P is generated from a covariance function.
One example of a covariance function is the Gaussian covariance: P_{ij} = \sigma^2 \exp\left(-\sum_k\frac{|\xi_{jk}-\xi_{ik}|^2}{2 \theta_k^2}\right), where we sum over the dimensions k and \xi are the input parameters. The hyperparameters \mu , \sigma and \theta can be estimated from a Maximum Likelihood Estimate (MLE).
Text derived from https://en.wikipedia.org/w/index.php?oldid=1369854665 under CC BY-SA 4.0; modifications have been made.