Reference: A classification model was fitted to the data medical data to try to predict whether an individual smoked (labeled as "Yes") or did not smoke (labeled as "No"). The following data attributes were added to the model, the attribute name is in parentheses:

• Age of individual (age)
• Sex at birth (sex)
• Body mass index (bmi)
• Number of children (children)
• Region of US (region)
• Medical charges (charges)

The training data had about 80% of the individuals as not smoking and the other 20% did smoke. The test data had about 78% of the individuals not smoking and 22% as smokers.

Below is the classification tree from the training data. charges = 160+3 charges = 32e+3 bmi >= 30