GIN · internal_only
Standard SplitFeb 17, 202690a92946cd714714b82f1d7d79f78cf0
Description
First baseline experiment. Train GIN on the internal_only FCG dataset using a standard stratified 70/15/15 train/val/test split to establish a performance reference for graph-based ransomware detection.
Conclusion
Strongest single-split baseline result: 98.2% accuracy with perfect malware recall (100%). Establishes GIN on internal_only as the top baseline. However, this result uses a standard random split where related ransomware variants can appear in both train and test sets.
Test Metrics
Accuracy
98.2%
F1 Macro
97.8%
F1 Malware
97.0%
Precision
94.1%
Recall
100.0%
AUROC
99.0%
Best Val Loss
0.0692
Training Time
883.9000s
Confusion Matrix
| Pred Benign | Pred Malware | |
|---|---|---|
| Actual Benign | 74 | 2 |
| Actual Malware | 0 | 32 |
Configuration
| Hidden Dim | 128 |
| Num Layers | 3 |
| Dropout | 0.5 |
| Batch Size | 4 |
| Learning Rate | 0.001 |
| Weight Decay | 0.0001 |
| Max Epochs | 200 |
| ES Patience | 20 |
| ES Min Epochs | 100 |
| LR Patience | 10 |
| LR Factor | 0.5 |
| Mixed Precision | Yes |
| Random Seed | 42 |
| Epochs Trained | 135 |