Table 4 The impact of \(\alpha\) and \(\beta\) on experimental results in the CICIDS2017 test dataset.
From: MEMCAIN: a memory-enhanced hybrid CNN-attention model for network anomaly detection
\(\alpha\) | \(\beta\) | ACC% | REC% | FPR% |
|---|---|---|---|---|
0.6 | 1.5 | 98.43 | 95.95 | 0.29 |
0.8 | 1.3 | 98.75 | 94.23 | 0.23 |
0.9 | 1.2 | 99.08 | 97.07 | 0.17 |
1.1 | 1.0 | 98.68 | 96.21 | 0.24 |
1.2 | 0.9 | 98.64 | 94.09 | 0.25 |
