A Network Data Reinforcement Method Based on the Multiclass Variational Autoencoder

Table 4

The metrics of ADMs in the end of epoch 2.

Dataset	Model	Accuracy (%)	Loss	Val_Accuracy (%)	Val_Loss
Raw data	ADM (MLP)	95.54	0.1054	94.67	0.3213
	ADM (CNN)	90.38	0.3417	95.01	0.1551
Reinforced data	ADM (MLP)	97.93	0.0422	97.11	0.0345
	ADM (CNN)	99.99	0.0017	91.17	0.8283