| # Engine Fault Detection Model | |
| ## Model Overview | |
| This model predicts engine condition (Active / Faulty) using sensor data. | |
| ## Algorithm | |
| Gradiant Boosting Classifier | |
| ## Training Details | |
| - Objective: Maximize Recall | |
| - Class Imbalance: Handled via class weights / scale_pos_weight | |
| - Outlier Treatment: Coolant temperature capped using training data thresholds | |
| ## Dataset | |
| - Source: ShanRaja/Data | |
| - Splits: Train / Validation / Test | |
| ## Metrics (Validation) | |
| - Recall: 0.98 | |
| - Precision: 0.64 | |
| - F1-score: 0.78 | |
| ## Intended Use | |
| Early fault detection to minimize false negatives. | |