Table 6 Performance comparison of proposed framework with prior works.
Method | Accuracy | Precision | Recall | F1-score |
|---|---|---|---|---|
SVM-based IDS | 89.2% | 86.5% | 85.7% | 86.1% |
LSTM IDS | 93.8% | 92.3% | 91.5% | 91.9% |
CNN + GANomaly (Kale et al., 2022) | 94.1% | 91.7% | 93.2% | 92.4% |
Transformer + VAE | 96.2% | 94.8% | 95.1% | 94.9% |
Proposed | 97.4% | 96.3% | 95.9% | 96.1% |
