Pick the winner
Once tuning has been performed, it's time to pick the optimal hyperparameters from the results and build the final model. Two helpers from tidymodels
come in handy:
The function select_best()
extracts the optimal hyperparameters from a tuning results tibble, and finalize_model()
plugs these results into the specification, replacing the placeholders.
It's your turn to try this using the results of the last exercise! The objects tune_spec
, tune_results
, and customers
are still loaded.
This exercise is part of the course
Machine Learning with Tree-Based Models in R
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Select the parameters that perform best
final_params <- ___
final_params