Classification

The classification table shows the practical results of using the network. For each case, the predicted response is Yes if that cases's predicted pseudo-probability is greater than 0.5. For each sample:

• Cells on the diagonal of the cross-classification of cases are correct predictions.
• Cells off the diagonal of the cross-classification of cases are incorrect predictions.

Of the cases used to create the model, 74 of the 124 people who previously defaulted are classified correctly. 347 of the 375 non-defaulters are classified correctly. Overall, 84.4% of the training cases are classified correctly, corresponding to the 15.6% incorrect shown in the model summary table. A better model should correctly identify a higher percentage of the cases.

Classifications based upon the cases used to create the model tend to be too "optimistic" in the sense that their classification rate is inflated. The holdout sample helps to validate the model; here 74.6% of these cases were correctly classified by the model. This suggests that, overall, your model is in fact correct about three out of four times.

Next