Get startedGet started for free

Finalizing the model

It is time to implement the results of your tuning work and impress the Human Resources team. You can finalize your model with the optimal penalty identified and see if it meets your expectations. Your results have been loaded, and the user-defined function class_evaluate() is available in your environment.

This exercise is part of the course

Feature Engineering in R

View Course

Exercise instructions

  • Select the optimal penalty for the Lasso.
  • Fit a final model using the optimal penalty.

Hands-on interactive exercise

Have a go at this exercise by completing this sample code.

# Select the optimal penalty for the Lasso
best_penalty <- ___(tune_output, metric = 'roc_auc', desc(penalty)) 
best_penalty

# Fit a final model using the optimal penalty
final_fit <- ___(workflow_lasso_tuned, best_penalty) %>%
  fit(data = train)

final_fit %>% tidy()

final_fit %>% augment(test) %>% class_evaluate(truth = Attrition, 
                                   estimate = .pred_class,
                                   .pred_Yes)
Edit and Run Code