Table 4 Performance evaluation using cross-validation with mean ± standard deviation.
From: Integrating AI in security information and event management for real time cyber defense
Dataset | Method | ACC (%) | SP (%) | SN (%) | F1 Score (%) | MCC |
|---|---|---|---|---|---|---|
Dataset 1 | BERT | 98.47 | 97.36 | 98.68 | 98.69 | 0.945 |
ID-CNNs | 98.60 | 97.82 | 98.74 | 98.75 | 0.950 | |
Hybrid Features (Before BTG) | 98.76 | 98.09 | 98.87 | 98.89 | 0.955 | |
Hybrid Features (After BTG) | 99.01 ± 0.08 | 98.20 | 99.16 | 99.17 ± 0.07 | 0.964 ± 0.005 | |
Dataset 2 | BERT | 97.18 | 96.90 | 97.32 | 97.33 | 0.938 |
ID-CNNs | 97.42 | 96.83 | 97.73 | 97.74 | 0.944 | |
Hybrid Features (Before BTG) | 97.55 | 96.90 | 97.89 | 97.90 | 0.946 | |
Hybrid Features (After BTG) | 97.63 ± 0.05 | 97.05 | 97.93 | 97.94 ± 0.04 | 0.948 ± 0.003 |
