Predict new values
A predictive model is one that predicts the outcomes of new, unseen data. Besides the numeric predictors, there are other useful columns in the dataset. The goal of this exercise is to predict the final rating grades of a chocolate tasting based on all other predictor variables that are available.
Loaded in your workspace is the regression tree specification that you created in the last exercise, chocolate_spec
, as well as the training and testing data (chocolate_train
and chocolate_test
).
This exercise is part of the course
Machine Learning with Tree-Based Models in R
Exercise instructions
- Use
model_spec
to train a regression treechocolate_model
that predictsfinal_grade
using all predictors in the data. - Predict
final_grade
values using thechocolate_test
tibble. - Add the results to the
chocolate_test
tibble.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Train the model
chocolate_model <- model_spec %>%
___
# Predict new data
predictions <- predict(___,
___) %>%
# Add the test set
bind_cols(___)
predictions