A Novel CGM Metric-Gradient and Combining Mean Sensor Glucose Enable to Improve the Prediction of Nocturnal Hypoglycemic Events in Patients with Diabetes

Table 3

The FPR and FNR in the periods of declining sensor glucose without a hypoglycemic event when the metric of MSG+gradient was adopted at the prediction horizon of 30 minutes.

Algorithm	LR	SVM	RF	LSTM
FPR (%)	9.32	12.20	7.15	7.69
FNR (%)	7.95	6.62	9.27	7.95

FPR: false positivity rate; FNR: false-negative rate; LR: logistic regression; SVM: support vector machine; RF: random forest; LSTM: long short-term memory.