Research Article
Few-Shot Learning-Based Network Intrusion Detection through an Enhanced Parallelized Triplet Network
Table 1
Label distribution of few-shot datasets based on the CICIDS-2017 dataset.
| Describe | Type | Train | Test |
| Normal traffic | Benign | 110 | 10000 |
| Known attack traffic | FTP-Patator | 10 | 1000 | SSH-Patator | 10 | 1000 | DoS Hulk | 10 | 1000 | DoS GoldenEye | 10 | 1000 | DoS Slowloris | 10 | 1000 | DoS Slowhttptest | 10 | 1000 | Web Attack-Brute Force | 10 | 1000 | Web Attack-XSS | 10 | 642 | Bot | 10 | 1000 | DDoS | 10 | 1000 | Port Scan | 10 | 1000 |
| Unknown attack traffic | Heartbleed | 0 | 11 | Infiltration | 0 | 36 | Web Attack-Sql Injection | 0 | 21 |
| ā | Total | 220 | 20710 |
|
|