A Novel MEGNet for Classification of High-Frequency Oscillations in Magnetoencephalography of Epileptic Patients

Table 2

The performance of various models was compared.

Method	Accuracy	Recall	Precision	F1-score
1D-CNN	1	1	1	1
CNN	0.89	0.89	0.90	0.89
DNN	0.92	0.92	0.92	0.92
RNN	0.88	0.88	0.88	0.89
CapsuleNet	0.86	0.85	0.85	0.86
MEGNet	0.94	0.94	0.95	0.94

The highest metrics were marked as bold.