← Back to All Projects 
Python RegressionFeature ImportanceReservoir Engineering 
![Screenshot full size]()
Machine Learning
Gas EUR Prediction
Python RegressionFeature ImportanceReservoir Engineering The Problem
Predicting Estimated Ultimate Recovery (EUR) for horizontal multistage fractured wells is critical for reservoir planning and investment decisions.
Approach
Built a regression model using geological, drilling, and completion attributes for 507 simulated wells.
Applied preprocessing, feature engineering, and model evaluation, plus feature importance analysis.
Results
Achieved high correlation with actual EUR values and identified the most influential well parameters for reservoir performance forecasting.
Key result: High correlation with actual EUR for 507 wells
Tools Used
PythonScikit-learnPandasSeaborn
GitHub - Coming Soon