Clues from networks: quantifying relational risk for credit risk evaluation of SMEs

Financial Innovation

Table 3 Discrimination performance of relational risk in relationship-specific networks

Model	Metric	BF	BF + CN	BF + DN	BF + SN
LR	AUC	0.715	0.722	0.729	0.725
	KS	0.408	0.414	0.431	0.417
	H	0.251	0.266	0.283	0.269
RF	AUC	0.725	0.738	0.741	0.735
	KS	0.426	0.444	0.450	0.439
	H	0.265	0.289	0.294	0.281
XGB	AUC	0.695	0.718	0.722	0.720
	KS	0.387	0.415	0.434	0.424
	H	0.219	0.253	0.257	0.245

Each metric is measured by the mean of 100 estimates obtained by repeating the outer ten-fold cross-validation procedure ten times. In Appendix 6, Tables 29 and 31 show the 95% and 90% confidence intervals of each mean, and Table 33 shows the mean of 50 estimates obtained by repeating the ten-fold cross-validation procedure five times