Rajesh Basnet completed Energy and Renewable: Machine Learning for the Power Grid

Built a leakage-free turbine yield predictor end to end: framed TEY as the target, used only the eight available sensor inputs, and excluded downstream CO/NOX emissions to avoid leakage. Validated with a chronological train/test split, compared a linear baseline to a tuned random forest, and reported honest held-out performance (about 1.5 MWh MAE, RMSE ~1.9, R²≈0.98) with a clear residual and monitoring read.

• Leakage-Free Feature Selection
• Honest Time-Based Evaluation
• Residual and Drift Interpretation
• Clear Non-Technical Communication