Subset Summary

The subset summary displays the order in which the predictors are added. Each additional predictor forms a new subset that contains the added predictor and previously added predictors; thus, subset 3 contains Years at current address, Debt to income ratio (x100), and Years with current employer.
- Subsets are ranked by the test data criterion, which is the negative of the average log-likelihood for the test data; note that the values in the Average Log-Likelihood column are for the training data. The subset with the lowest test data criterion is ranked first.
- The predictors in the rank 1 subset, Years with current employer, Debt to income ratio (x100), Years at current address, Credit card debt in thousands, and Level of education, are selected by the procedure and used in the Naive Bayes model.