This table represents the result of Preprocessing and Feature Engineering.
Random Forest Model Performance
Evaluation metrics based on internal model validation (Best Ratio: -)
Accuracy
-
Precision
-
Recall
-
F1 Score
-
Confusion Matrix
Predicted Class
Actual Class
Normal (0)
Attack (1)
Normal (0)
-
TN
-
FP
Attack (1)
-
FN
-
TP
Feature Importance
Cannot calculate feature importance
Data contains only one class (all Attacks or all Normal)
Class Distribution
ROC Curve
ROC Curve not available (Single class data)
Overall Performance Balance (Radar Chart)
Model Performance by Split Ratio (80:20 vs 70:30 vs 60:40)
Comparison of Accuracy, Precision, Recall, and F1-Score across different training/testing split ratios.
These metrics are calculated by training a Random Forest Classifier on the best performing split ratio.
The Rule-Based Detection results are used as the "Ground Truth" labels for this evaluation.