Reliable and Adaptive Probabilistic Forecasting for Event-Driven Water-Quality Time Series Using a Gated Hybrid–Mixture Density Network
Nadir Ehmimed, Mohamed Yassin Chkouri, Abdellah Touhafi · 2025
Read Reliable and Adaptive Probabilistic Forecasting for Event-Driven Water-Quality Time Series Using a Gated Hybrid–Mixture Density Network by Nadir Ehmimed, Mohamed Yassin Chkouri, Abdellah Touhafi (2025) 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-12737238.txt.
Description
Nadir Ehmimed, Mohamed Yassin Chkouri, and Abdellah Touhafi build a water quality forecaster that widens its error bands during storms in a 2025 paper. River quality swings hard during events, and standard models state false confidence then. Their network carries a gate that reads an event precursor signal. The gate stretches or narrows forecast bands to fit the moment.
Tests on fake and real data show sound calibration across calm and wild spells. Bands widened to catch extreme events where rival models missed. The authors warn that common average scores can hide overconfidence in rare events. Judging calibration first, they argue, builds forecasts that managers can trust for risk calls.
Catalog record
| CREATOR | Nadir Ehmimed, Mohamed Yassin Chkouri, Abdellah Touhafi |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 70 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |