 |
|
This DOE algorithm uses the Cross Validation method based on the Kriging algorithm used for the response surfaces.
This method estimates the error of the Kriging model and then, according to the standard error distribution, it chooses a good new set of input values in order to make the Kriging model more reliable.
So it distributes the designs uniformly in the variables space according to the residuals.
Three parameters should be defined:
Number of Designs a positive integer number between 1 and the 100.
K-nearest Designs is the number of neighbor points to consider. The maximum number of neighbors is 100 even if it is also limited by the real number of designs present in the database.
Random Generator Seed a positive integer number, used for sequence repeatability. If the seed value is 0, the sequence is automatically seeded to a value based on the current time.
