Gespeichert in:
| 1. Verfasser: | |
|---|---|
| Format: | Recurso digital |
| Sprache: | |
| Veröffentlicht: |
Zenodo
2025
|
| Online-Zugang: | https://doi.org/10.5281/zenodo.15247669 |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Inhaltsangabe:
- <p>This upload contains the final model and output files developed as part of the second task (“Part 2 – Data Management”) of the Data Stewardship UE 2025S course at TU Wien.</p> <p>The goal of the project was to implement a machine learning pipeline to analyze and predict CO₂ emissions based on real-world socio-economic indicators. A Random Forest Regressor was used after data cleaning, preprocessing, and splitting the data into training, validation, and test subsets using the DBRepo API.</p> <p>⚠️ Due to technical limitations, only partial subsets of the dataset were uploaded to DBRepo. This may have impacted the predictive performance of the trained model.</p> <p>The full code, evaluation metrics, and visualizations (including RMSE, R², scatter plot comparisons, and feature importance) are available in the GitHub repository.</p> <p> GitHub: <a href="https://github.com/KhasrurRahman/Data-Stewardship-UE-2025S---Data-Management-part--2-">https://github.com/KhasrurRahman/Data-Stewardship-UE-2025S—Data-Management-part–2-</a></p> <p> Includes:</p> <ul> <li> <blockquote>Final trained Random Forest model</blockquote> </li> <li> <blockquote>Output prediction CSVs</blockquote> </li> <li> <blockquote>Feature importance visualization</blockquote> </li> <li> <blockquote>Requirements.txt file</blockquote> </li> <li> <blockquote>Data Management Plan (PDF)</blockquote> </li> </ul> <p> </p>