Configuring Kriging Model Technique Options

Kriging approximation is an interpolation technique. Kriging approximations are extremely flexible because you can choose from a wide range of correlation functions to build the meta model. In addition, depending on the correlation function that you choose, the meta model can either “honor the data,” providing an exact interpolation of the data, or “smooth the data,” providing an inexact interpolation.

For more information, see Kriging Model.

  1. Double-click the Approximation component icon .

    The Approximation Component Editor appears.

  2. From the Approximation Component Editor, click the Technique Options tab.

  3. Select the Fit Type:

    Option Description
    Anisotropic Select this type if the independent variables represent different physical measures (e.g., time, distance, velocity, etc.) or when the independent variables have different scales. Anisotropic fit is the general case for ordinary Kriging when every independent variable behaves differently.
    Isotropic Select this type if all the independent variables behave similarly. Isight handles all θ k values as if they are identical. Typically, the Isotropic fit is faster than the Anisotropic fit because Isight searches for only one optimum theta value.

  4. Select the Correlation Function. The correlation functions interpolate the data points exactly.

    The following options are available:

    Option Description
    Gaussian You can use the Gaussian correlation function for approximating smooth functions. However, it produces a poor fit when sampling points are too close.
    Exponential If the sample points are close, use the Exponential correlation function.
    Matern Linear You can use the Matern Linear correlation function if the Gaussian and Matern Cubic correlation functions produced an unacceptable fit. The Matern Linear correlation is more robust, but less accurate, than the Matern Cubic correlation function.
    Matern Cubic You can use the Matern Cubic correlation function if the Gaussian correlation function produced an unacceptable fit. Typically, the Matern Cubic correlation function is more accurate than the Matern Linear correlation function.

  5. Enter the Filter Distance.

    Occasionally, when points are clustered together the matrices used in fitting the Kriging model become ill-conditioned resulting in a poor fit. You can filter points from the sample based on distance to avoid a poor fit. All points that are closer than the Filter Distance are removed from the sample set before fitting. Isight uses other numerical techniques internally to improve the performance and robustness of the approximation.

    Note: When the number of points is large, Kriging automatically filters out points that are clustered until only 500 points remain even if the Filter Distance value is zero. Isight builds the Kriging model using these 500 points.

  6. Enter the Maximum Iterations to Fit.

    Isight uses an iterative procedure to fit a Kriging model with the maximum likelihood estimate. Isight limits the maximum number of iterations used to fit a model for each response based on this value.

  7. Click OK to save your changes and to close the Approximation Component Editor.