Table 12 Effects of different models, sample sizes, and datasets on the test accuracy.
From: Multimodal fusion based few-shot network intrusion detection system
Model and sample | CICIDS2017 | CICIDS2018 |
|---|---|---|
SAF + S-1 | 69.8% | 89.8% |
SAF + T-1 | 78.0% | 90.2% |
LF + S-1 (baseline) | 74.1% | 71.3% |
