What Noah demonstrated

Built a leakage-aware turbine yield predictor end to end: framed TEY as the target, excluded downstream CO/NOX emissions, trained a LinearRegression baseline and a RandomForest model, and validated on a locked-away later-year holdout. The final model achieved about 1.5 MWh MAE and 0.98 R² on unseen data, with a model card that clearly states intended use, limitations, and monitoring guidance.

• Leakage-Free Feature Selection
• Time-Based Holdout Evaluation
• Honest Model Comparison
• Plain-Language Model Communication