Table 2 Precision, Recall, and FNR for each attack type using FedGATSage on NF-ToN-IoT and CIC-ToN-IoT datasets.
From: Graph-based federated learning approach for intrusion detection in IoT networks
Attack Type | NF-ToN-IoT | CIC-ToN-IoT | ||||
|---|---|---|---|---|---|---|
Precision | Recall | FNR | Precision | Recall | FNR | |
Benign | 0.9986 | 0.9944 | 0.0056 | 1.0000 | 0.9785 | 0.0215 |
Backdoor | 0.9451 | 0.9942 | 0.0058 | 0.9816 | 1.0000 | 0.0000 |
DDoS | 0.8939 | 0.5322 | 0.4678 | 0.8939 | 0.5322 | 0.4678 |
DoS | 0.3896 | 1.0000 | 0.0000 | 0.3896 | 1.0000 | 0.0000 |
Injection | 0.9760 | 0.3412 | 0.6588 | 0.7266 | 0.5146 | 0.4854 |
Password | 0.3890 | 0.6226 | 0.3774 | 0.5268 | 0.7296 | 0.2704 |
Scanning | 0.1050 | 0.8029 | 0.1971 | 0.7634 | 0.9302 | 0.0698 |
XSS | 0.4197 | 0.9990 | 0.0010 | 0.9186 | 0.6612 | 0.3388 |
