Self-directed project
Energy and Renewable: Machine Learning for the Power Grid
Junior Machine Learning Engineer · AI / ML · September 2026
A ProoV case study · educational project, not employment
Built a leakage-aware gas-turbine TEY predictor end to end: selected the eight legitimate sensor features, excluded downstream CO/NOx emissions, and validated on a locked later-period test split. Compared a LinearRegression baseline with a RandomForest model and showed the forest improved held-out MAE from 2.90 to 1.53 MWh with R² rising to 0.983. Also authored a concise model card covering intended use, limitations, and drift monitoring.
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 documented why they were removed
- Used an earlier-period train / later-period test split instead of a random shuffle
- Compared a LinearRegression baseline to a tuned RandomForest and showed a meaningful MAE improvement
- Included a practical monitoring idea based on input-range drift and rolling error
- Leakage-Free Feature Selection
- Time-Based Model Validation
The work I submitted17 tasks
9 tasks · 5.0k characters · Python, JavaScript · 8 written answers · 408 words
- Problem Frame170 characters
- Leakage Flags113 characters
- Train Validate BaselinePython · 1.4k characters
- Train Validate Forest943 characters
- Train Validate124 characters
- Residual Read117 characters
- Drift Monitor94 characters
- Model Card1.3k characters
- Teach BackJavaScript · 789 characters
- Problem Frame13 words
- Leakage Flags1 word
- Train Validate Baseline141 words
- Train Validate Forest76 words
- Train Validate1 word
- Residual Read8 words
- Drift Monitor1 word
- Model Card167 words
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.