Skip to main content

Table 5 Performance of the model. To evaluate the performance of the model, we use its predictive power for data outside the model. 54,000 records that were not used in training were entered into the model and their probability was calculated and compared with the values of the table. The degree of sensitivity and diagnosis of the model were 0.84.05% and 89.23%, respectively

From: A dynamic credit risk assessment model with data mining techniques: evidence from Iranian banks

Real estimate observation Medium risk customers High risk customers Total
P(Y) < =0.37 638 58 696
P(Y) > 0.37 77 307 384
Total 715 365 1080
Correct 638 307 945
Correct percent 89.23% 84.05% 87.5%
Incorrect percent 10.76% 84.1% 12.5%