Batch 5 · Threat Console
JBJ IDS Console
Network Intrusion Detection · CSE-CIC-IDS2018 · 14 attack classes
Analyse network traffic
Upload a CICFlowMeter CSV with the same columns as CSE-CIC-IDS2018. A Label column is optional; if it is present, the report also scores the model against it.
Deployed model
- Model
- Random Forest + class bias
- Trained
- 2026-10-10T14:31:03+00:00
- Test macro-F1
- 0.8814
- Test accuracy
- 0.9260
- Features used
- 40
- Anomaly detector
- Autoencoder (benign-only)
The supervised model names the attack type. The autoencoder flags flows that look unlike normal traffic, even when they match no known class.