Intrusion Detection on CICIDS 2017 (Accuracy, Precision, Recall, F1)
99.97AccuracyRF
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| RF2026.06 | 99.97 | — | — | — | |
| Basnet et al.Attack Types=Network-based2026.06 | 99.95 | 100 | 99.8 | 99.8 | |
| KAN-LSTM2026.03 | 99.28 | 99.06 | 99.3 | 99.66 | |
| Deep LearningYear=2025, Reference=LaxmiLydia et al. [17]2026.03 | 98.77 | 98.77 | 98.77 | 98.76 | |
| ConvKAN2026.03 | 98.55 | 91.48 | 99.69 | 95.41 | |
| MLPNumber of layers=52026.03 | 98.43 | 77.03 | 77.07 | 77.04 | |
| KANsNumber of layers=52026.03 | 98.2 | 77.03 | 77.07 | 77.04 | |
| KANsNumber of layers=22026.03 | 97.69 | 77.03 | 76.78 | 76.89 | |
| MLPNumber of layers=22026.03 | 97.61 | 77.03 | 78.1 | 77.55 | |
| 1D-CNNAuthor=Kim et al., Year=20172026.06 | 97.1 | — | — | — | |
| LSTM2026.03 | 86.05 | 86.323 | 85.68 | 85.47 | |
| CNN2026.03 | 82.75 | 82.75 | 40.5 | 43.89 | |
| CNN-LSTM2026.06 | — | — | — | 97.8 | |
| MDE-IDS (LGB)backbone=LightGBM combined2026.06 | — | — | — | 99.89 |