Table 9 37-Class classification accuracy on multiple subsets.
From: Graph attention and Kolmogorov–Arnold network based smart grids intrusion detection
Subset | GCN-FC (%) | GAT-FC (%) | KAN-MLP (%) | GraphKAN (%) |
|---|---|---|---|---|
1 | 94.46 | 96.27 | 97.51 | 98.48 |
2 | 97.95 | 96.61 | 98.25 | 98.55 |
3 | 98.31 | 98.87 | 98.20 | 99.08 |
4 | 99.23 | 99.99 | 97.93 | 98.56 |
5 | 97.73 | 97.13 | 98.25 | 98.84 |
6 | 97.97 | 97.58 | 98.05 | 98.78 |
7 | 97.79 | 98.54 | 98.31 | 99.21 |
8 | 98.93 | 99.11 | 98.77 | 99.11 |
9 | 98.06 | 97.42 | 98.34 | 98.79 |
10 | 98.36 | 98.64 | 98.47 | 99.21 |
11 | 98.99 | 99.04 | 98.53 | 99.40 |
12 | 98.95 | 97.97 | 98.65 | 98.95 |
13 | 98.37 | 98.15 | 98.52 | 99.11 |
14 | 98.28 | 95.82 | 98.65 | 98.59 |
15 | 98.90 | 98.87 | 98.69 | 99.44 |
