netgoat-ai/koda-2

Model data

Total params: 6,382 (24.93 KB)

Trainable params: 5,982 (23.37 KB)

Non-trainable params: 400 (1.56 KB)

---

Benchmark results

netgoat-ai/SynthDDoS

  Accuracy:  0.9822 
  Precision: 0.9655  (of flagged traffic, how much was real attack?)
  Recall:    1.0000  (of all attacks, how many did we catch?)

  Confusion Matrix:
                   Predicted Benign | Predicted Attack
  Actual Benign:   96490            | 3563
  Actual Attack:   0                | 99755

CIC-DDoS2019

  Accuracy:  0.7747
  Precision: 0.7744
  Recall:    0.9998

  Confusion Matrix:
                   Predicted Benign | Predicted Attack
  Actual Benign:   141              | 19425
  Actual Attack:   13               | 66696