Table 3 Comparison of structure modified models and PCA models.
| Â | Accuracy | Precision | Recall | F1-Score | Time (ms) |
|---|---|---|---|---|---|
Basic neural network with PCA | 98.6% | 98.6% | 98.6% | 98.6% | 1.18 |
Basic neural network with ResNet | 95.2% | 95.2% | 95.6% | 95.4% | 1.23 |
Deep neural network with PCA | 97.6% | 97.6% | 97.8% | 97.7% | 1.03 |
Deep neural network with ResNet | 97.1% | 97.1% | 97.2% | 97.1% | 1.82 |
Wide neural network with PCA | 99.0% | 99.0% | 99.0% | 99.0% | 2.87 |
Wide neural network with ResNet | 67.6% | 68.0% | 71.9% | 69.9% | 7.36 |
CNN | 98.6% | 98.6% | 98.6% | 98.6% | 2.30 |
CNN with PCA | 98.6% | 98.6% | 98.6% | 98.6% | 0.96 |
RNN | 95.2% | 95.2% | 95.4% | 95.3% | 4.37 |
RNN with PCA | 99.1% | 99.1% | 99.0% | 99.0% | 2.84 |
FWCW-ResNet35 | 87.1% | 88.2% | 89.0% | 87.2% | – |
5G-enabled36 | 94.6% | 95.1% | 96.7% | 94.6% | – |
CNN-LSTM37 | 92.0% | 92.3% | 92.3% | 91.7% | – |
