Cross Validation

If no test data set is available, an n-fold cross-validation can be carried out on the training data set. To do so, proceed as follows.

Cross-validation

• Select Model > Cross Validation on Training Set.

  

• In the Cross Validation window, enter the number of data sets you want to use for cross-validation.

• Click OK.

  The data set is split into n equal parts, according to the specified number of cross validations.

For each output, n plots are displayed. Each plot is the result of an individual model training on n-1 data sets and the prediction on the data set not used for modeling.