Table 4.

Metrics for both the 1-year and 1-month model. Values shown are the average of the 10 cross-validations. SS: standard scaling; SMOTE: Synthetic Minority Oversampling Technique; AUC: area under the receiver operating curve; LR+: positive likelihood ratio; LR−: negative likelihood ratio.

	Model Type	Scaling	Pre-Processing	AUC (Test)	AUC (Train)	Accuracy	Sensitivity	Specificity	LR+	LR−
1-month	Random forest	Not used	Not used	0.730	0.758	0.68	0.63	0.7	3.2	0.53
1-year	Random forest	SS	SMOTE	0.713	0.751	0.71	0.61	0.77	2.8	0.5