Tune Model Hyperparameters Module: tune and cross validate at the same time

Question

suppose I'm using the following configuration with the Tune Model Hyperparameters module with a Boosted Decision Tree Regression:

Does this configuration let me tune the hyperparameters so that I'll have the best Coefficient of Determination, and at the same time it guarantees the lowest cross-validation average error? If so, does anyone knows what the Tune Model Hyperparameters module do using this configuration in more details?

Thank you.

Answer 1

No need of splitting the dataset as train/test is you going to use cross validation to measure the accuracy of the model. No validation data-set needed as a input for the Tune model hyperparameter too.

Here's a simple configuration built to measure the accuracy of a binary classification. See the properties of the Partition & Sample module has adjusted for 10-fold cross validation with a random split.

The evaluation results, indicates the parameters that produced the best models, and their accuracy. The accuracy metrics are calculated from the cross-validation pass, and may vary slightly depending on how many folds you selected.