Skip to main content

View full-text article in PMC

. 2022 Apr 6;10(4):e29290. doi: 10.2196/29290

Table 3.

Result indicators of each model.

Model	Accuracy	F1 score	Log-loss
Cross-FGCNN^a	0.9621	0.9621	0.8356
Decision tree	0.7448	0.7439	6.4533
10-layer ANN^b	0.9121	0.9115	1.9071
ML-KNN^c	0.9075	0.9076	2.7211
Hypergraph clustering	0.8816	0.8814	3.8436
Bayesian	0.7816	0.7815	4.5555
SVM^d	0.8992	0.8989	3.2289
Deep & cross network	0.7992	0.7997	3.1602
FGCNN	0.9390	0.9390	1.2820
DNN^e	0.7220	0.6804	3.9439

^aFGCNN: feature generation by convolution neural network.

^bANN: artificial neural network.

^cML-KNN: multilabel K nearest neighbor.

^dSVM: support vector machine.

^eDNN: deep neural network.