Table 10 the accuracy of the ablation experiments on ADGCN.
From: Few-shot traffic classification based on autoencoder and deep graph convolutional networks
 | USTC-TFC2016 | MALWARE | NORMAL | ISCX-VPN-NonVPN-2016 | VPN | NO-VPN |
---|---|---|---|---|---|---|
ADGCN-AE | 47.68 | 53.12 | 56.82 | 56.77 | 86.87 | 57.01 |
ADGCN-Heat Kernel | 70.85 | 87.09 | 81.83 | 79.33 | 94.52 | 77.67 |
ADGCN-GCNII | 91.55 | 80.99 | 93.49 | 88.08 | 88.01 | 77.33 |
ADGCN | 96.2 | 98.6 | 98.8 | 94.33 | 97.67 | 96.67 |