Table 3.

Comparative analysis between classification results before and after implementing VAE. The best results are highlighted in bold.

Classifier	Metrics	Before VAE	After VAE	After DVAE
CNN	Loss	0.9449	0.5545	0.6932
CNN	Accuracy	0.6876	0.7365	0.5002
CNN	Precision	0.6805	0.7369	0.2502
CNN	Recall or Sensitivity	0.6984	0.7365	0.5002
CNN	F1 Score	0.6868	0.7363	0.3335
CNN	Specificity	0.69	0.74	0.50
CNN	Cohen’s Kappa	0.3749	0.4730	0.000
CNN	ROC AUC	0.7205	0.7969	0.5000