48-hour Data Analysis Competition
UD DataQuest 2026
A streamflow-forecasting project exploring model comparison, storm-event features, and weighted ensembles.
Best Modeling
PythonXGBoostGRUscikit-learn
Sanitized project visual placeholder
Competition context
UD DataQuest 2026 was a 48-hour undergraduate data-analysis competition. This project received the Best Modeling award.
Engineering focus
- Streamflow forecasting
- Model comparison across LR, GRU, Ridge, and XGBoost approaches
- Weighted ensembles
- KGE, R², MSE, and NSE evaluation metrics
- NOAA storm-event features
Publication note
This scaffold stays intentionally conservative about final-model claims. The detailed write-up should be expanded only from verified competition materials.