Table 4.

Machine learning model performance.

Factor	Random forest	Logistic regression	Gradient boosting	Naive Bayes
True positives (TP)	60	55	55	44
True Negatives (TN)	20	16	22	18
False Positives (FP)	4	8	2	6
False Negatives (FN)	6	11	11	22
Accuracy	0.89	0.79	0.86	0.69
Precision	0.94	0.87	0.96	0.88
Sensitivity (Recall)	0.91	0.83	0.83	0.67
Specificity	0.83	0.67	0.92	0.75
F1 Score	0.92	0.85	0.89	0.76
AUC-ROC	0.91	0.82	0.89	0.74
AUC-PR	0.91	0.88	0.91	0.84

AUC, Area Under the Curve; ROC, Receiver Operating Characteristic; PR, Precision-Recall.