The response (outcome) variable is whether the customer defaulted on their debt. The

predictor is the average balance. The following R output is the fitted logistic regression model.

Coefficients:

EstimateStd. Errorz valuePr(>|z|)

(Intercept)-10.870.5170250-10.08<2e-16 ***

balance 0.0017440.000249411.12<2e-16 ***

---

From the output, will a customer with the average debt of $8000 default?

Cannot determine since each customer is different.

The customer is very unlikely to default.

The customer is very likely to default.

The customer will default.

Tree-based models rely on which two key ideas?

recursive partitioning and pruning

gini index and recursive partitioning

pruning and gini index

classification and pruning