ComenzarEmpieza gratis

Compare to a single classifier

You learned that boosted trees are among the best machine learning algorithms available. To underpin that, you will now compare the AUC of a boosted model to a simple decision tree classifier.

In this exercise, you will use the whole machine-learning pipeline from specification to evaluation.

The training data customers_train is still pre-loaded.

Este ejercicio forma parte del curso

Machine Learning with Tree-Based Models in R

Ver curso

Ejercicio interactivo práctico

Prueba este ejercicio completando el código de muestra.

set.seed(100)

# Specify, fit, predict, and combine with training data
predictions <- boost_tree() %>%
  set_mode("___") %>%
  set_engine("___") %>% 
  fit(still_customer ~ ., data = ___) %>%
  predict(new_data = ___, type = "___") %>% 
  bind_cols(___)

# Calculate AUC
roc_auc(___, 
        truth = ___, 
        estimate = ___)
Editar y ejecutar código