Skip to main content

Table 3 Performance of the recommend shallow neural network classifier models in validation dataset

From: Early warning of citric acid overdose and timely adjustment of regional citrate anticoagulation based on machine learning methods

Labels

AUC

Precision (%)

Recall (%)

F1-score (%)

Accuracy (%)

“0”: < 0.25 mmol/L

0.8638

80.00

80.00

80.00

80.00

“1”: 0.25–0.35 mmol/L

0.8086

80.75

80.50

80.46

80.50

“2”: 0.35–0.5 mmol/L

0.8466

80.37

80.37

80.37

80.36

“3”: > 0.5 mmol/L

0.7919

78.45

78.00

78.90

78.32