LoslegenKostenlos loslegen

Fit the folds

Now that you split your data into folds, it's time to use them for model training and calculating the out-of-sample error of every single model. This way, you gain a balanced estimation of the performance of your model specification because you evaluated it out-of-sample several times.

Provided in your workspace is chocolate_folds, which you created in the last exercise (10 folds of the chocolate training set).

Diese Übung ist Teil des Kurses

Machine Learning with Tree-Based Models in R

Kurs anzeigen

Anleitung zur Übung

  • Show that you can still do it: create tree_spec, a regression tree specification using an "rpart" engine.
  • Use fit_resamples() to fit your folds to tree_spec, modeling final_grade using all other predictors and evaluating with both MAE and RMSE.

Interaktive Übung

Vervollständige den Beispielcode, um diese Übung erfolgreich abzuschließen.

# Create a specification
tree_spec <- ___ %>%
    ___
    ___

# Fit all folds to the specification
fits_cv <- ___(tree_spec,
               ___,
               resamples = ___,
               metrics = ___)

fits_cv
Code bearbeiten und ausführen