Table 5.

The 10-fold cross-validation model performance in the development cohort for the short-term reversibility of renal function recovery.

	Logistic regression	XGBoost	Bayesian networks	Random forest	SVM
Accuracy	0.6728 ± 0.02	0.6445 ± 0.02	0.648 ± 0.03	0.6529 ± 0.03	0.6190 ± 0.03
Sensitivity	0.4694 ± 0.04	0.3939 ± 0.04	0.4489 ± 0.05	0.3762 ± 0.06	0.2852 ± 0.05
Specificity	0.8761 ± 0.03	0.8951 ± 0.02	0.8470 ± 0.02	0.9296 ± 0.02	0.9529 ± 0.02
AU-ROC	0.7669 ± 0.03	0.7328 ± 0.03	0.7477 ± 0.03	0.7683 ± 0.03	0.7187 ± 0.03