Table 2.

Methods performance comparison.

Models	Accuracy	Precision @0.8 recall	F1 score	AUPRCa	AUROCb
LSTM-DNN	93.04c	89.02c	84.30c	91.18c	96.15c
GRU-DNN	92.80	88.04	83.92	90.65	95.77
DNN	92.49	86.64	83.17	90.10	95.49
Logistic regression	90.77	81.86	80.03	86.47	93.96
Random forest	90.95	78.39	76.78	86.86	94.17

^aAUPRC: area under the precision–recall curve.

^bAUROC: area under the receiver operating characteristic curve.

^cNumbers for the best method.