← Back to The IndexFinal ReviewGroup 8
Final Review

Review Nº 08

Predicting the Presence of Data Centers in European NUTS3 Regions
AuthorsMohammad Mahdi Molla Jafari, Mohammadreza Goudarzi, Asma Zeynalpoor, Negin Alsadat Hosseini Gorji, Armin Akbari
25/30
Score
The submission gives a coherent classification analysis and clear main findings, but it lacks methodological depth, reproducibility details, and robust validation evidence.
Data Centres · Final Review

The Pros

1 Item
+
Clear binary classification framing; uses both statsmodels Logistic Regression and Random Forest; reports understandable performance metrics; identifies GDP, internet activity, population, and network quality as key predictors; includes policy interpretation and acknowledges that binary presence ignores facility scale.

The Cons

1 Item
−
Preprocessing and hyperparameter tuning are described too briefly; no cross-validation results are presented; feature definitions such as “internet usage” and “temperature” are not sufficiently precise; logistic regression inference is summarized without coefficient table; the report does not deeply discuss data-center collection quality, missingness, or leakage prevention.
↩ Back to The Index
Final Review · Group 8The IndexData Centres · 2026