metadata
language: en
tags:
- machine-learning
- classification
- predictive-maintenance
- engine-condition
- scikit-learn
datasets:
- dhani10/engine-condition-dataset
metrics:
- accuracy
- f1
- precision
- recall
- roc_auc
Engine Condition Prediction Model
This model predicts engine condition (Normal=0 / Faulty=1) from sensor signals for predictive maintenance.
Model Details
- Winning Model: Random Forest
- Training Data: dhani10/engine-condition-dataset
- Input Features: ['Engine rpm', 'Lub oil pressure', 'Fuel pressure', 'Coolant pressure', 'lub oil temp', 'Coolant temp']
- Target: Engine Condition (0=Normal, 1=Faulty)
- Training Samples: 15628
- Test Samples: 3907
- Registered: 2025-11-07T07:50:04.426743+00:00
Cross-Validation
- Best CV Score: 0.7724
Test Set Performance
- Accuracy: 0.6596
- F1-Score: 0.7684
- Precision: 0.6728
- Recall: 0.8957
- ROC-AUC: 0.6992
Best Hyperparameters
{
"classifier__max_depth": 5,
"classifier__n_estimators": 100
}