Self-directed project
Energy and Renewable: Machine Learning for the Power Grid
Junior Machine Learning Engineer · AI / ML · July 2026
A ProoV case study · educational project, not employment
Built a turbine-yield predictor end to end using leakage-free sensor inputs, a LinearRegression baseline, and a tuned RandomForest evaluated on later held-out data. Reported honest held-out performance (about 1.5 MWh MAE, 1.9 RMSE, R²≈0.98), checked residual behavior, and documented the model’s intended use, limits, and monitoring plan in a concise model card.
Graded against
- Build and validate the model correctly35%
- Optimize and evaluate honestly25%
- Write the Model Card (clarity + honesty)25%
- Explain it to a non-expert15%
Passed · pass mark 60/100
What stood out6
- Correctly excluded CO and NOx as downstream leakage features and explained why they are unavailable at prediction time.
- Used a genuine earlier-vs-later split instead of a random shuffle, which matches the deployment setting.
- Compared a linear baseline against a RandomForest and showed a clear held-out improvement in MAE and R².
- Read the residuals and identified a heteroscedastic pattern plus a realistic high-temperature failure mode.
- Leakage-Free Feature Selection
- Time-Based Holdout Evaluation
The work I submitted20 tasks
12 tasks · 13 lines · Python, JavaScript · 8 written answers · 442 words
- Problem Frame184 characters
- Leakage Flags114 characters
- Train Validate BaselinePython · 1.4k characters
- Train Validate Forest944 characters
- Train Validate124 characters
- Tune Evaluate SetupPython · 1.5k characters
- Tune Evaluate ResidualsPython · 2 lines
- Tune Evaluate132 characters
- Residual Read335 characters
- Drift Monitor95 characters
- Model Card1.5k characters
- Teach BackJavaScript · 1.1k characters
- Problem Frame14 words
- Leakage Flags1 word
- Train Validate Baseline140 words
- Train Validate Forest77 words
- Train Validate1 word
- Tune Evaluate Setup110 words
- Tune Evaluate Residuals98 words
- Tune Evaluate1 word
Summarised on purpose — the submitted code and writing stay private so this page cannot be reused as an answer key. The full submission sits behind the verified certificate.