Table 6.

Performance comparison of tested classifiers for patient risk identification system (Acc for accuracy, Pre for precision, and F1 for F1-score).

Classifier	Acc	Pre	Recall	F1	Mean absolute error (MAE)	Root mean square error (RMSE)
Random forest	83	75	83	77	0.172	0.42
Logistic regression	88	70	84	76	0.163	0.404
Decision tree	74	74	74	74	0.257	0.507
K-nearest neighbor	82	73	82	76	0.18	0.425
Support vector machine	84	70	84	76	0.163	0.404
XGBoost	79	73	79	76	0.209	0.457