📊 Model Comparison
| Model | Accuracy (Test) | Recall Fail (0) | Recall Success (1) |
|---|---|---|---|
| Logistic Regression | 0.75 | ≈0.40 | ≈0.85 |
| Decision Tree | 0.74 | 0.55 | 0.82 |
| SVM (Optimized) | 0.67 | 0.36 | 0.79 |
| KNN (Optimized) | 0.69 | 0.45 | 0.79 |
| 🌟 Random Forest (Optimized) | 0.95 | 0.9373 | 0.9373 |
Preliminary Conclusion
The Random Forest (optimized) shows the best performance with the highest balance between detecting failures and overall accuracy. It is currently the most reliable model for this classification problem.