Sanidhya Verma completed Energy and Renewable: Machine Learning for the Power Grid

Built a turbine-yield predictor end to end: framed TEY as the target, removed downstream CO/NOx leakage features, trained a LinearRegression baseline and a RandomForest model, and validated on a locked-away later time period. Reported held-out MAE, RMSE, and R², checked residual behavior, and documented practical limits and drift monitoring in a model card written for operational use.

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