EAPO adapts wildfire models to new environments via k-nearest neighbor data retrieval and hybrid fine-tuning that emphasizes rare extreme events, achieving ROC-AUC 0.7310 on real data.
Miller, Christopher B
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Environment-Adaptive Preference Optimization for Wildfire Prediction
EAPO adapts wildfire models to new environments via k-nearest neighbor data retrieval and hybrid fine-tuning that emphasizes rare extreme events, achieving ROC-AUC 0.7310 on real data.