Data driven water quality assessment using machine learning and synthetic data generation
Naga Durga Satya Siva Kiran Relangi, D. Venkata Naga Raju, P. Venkata Rama Raju, M. Vishnu Vardhana Rao, Krishna Kant Pandey, Vinod Yadav · 2026
Read Data driven water quality assessment using machine learning and synthetic data generation by Naga Durga Satya Siva Kiran Relangi, D. Venkata Naga Raju, P. Venkata Rama Raju, M. Vishnu Vardhana Rao, Krishna Kant Pandey, Vinod Yadav (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13478298.txt.
Description
Relangi and colleagues tackled a data problem in 2026: water quality records are too small and lopsided to train good machine learning classifiers for judging whether water is drinkable. They generated synthetic records by oversampling, using a method called SMOTE and a rival approach using generative networks, then checked how closely the fake data matched the real data. Several classifiers were trained on each synthetic set and tested for accuracy.
The SMOTE drinking-water set matched real data closely, with a discrepancy score of 0.0067 where lower means closer to real. Gradient boosting and XGBoost classifiers trained on it hit 99.47 percent test accuracy in the reported experiments. The generative-network data scored worse on closeness despite some strong classifier results. On a second water dataset both methods produced close matches, leading the authors to favor oversampling for dependable synthetic training data.
Catalog record
| CREATOR | Naga Durga Satya Siva Kiran Relangi, D. Venkata Naga Raju, P. Venkata Rama Raju, M. Vishnu Vardhana Rao, Krishna Kant Pandey, Vinod Yadav |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 95 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |