Extreme-value forest fire prediction A study of the Loss Function in an Ordinality Scheme

2026-01-06

arXiv.org

AI Summary

This study addresses the challenge of predicting extreme wildfires, whose spatial distribution and intensity exhibit severe imbalance, by proposing the first ordinal classification framework tailored for operational wildfire risk decision-making in France. The approach integrates intensity ordering, data imbalance, and seasonal risk through ordinal-aware loss functions—specifically WKLoss and a newly introduced Truncated Discrete Exponential Generalized Pareto Distribution (TDeGPD)—aligned with fire intensity levels. Large-scale benchmarking on real-world data across multiple neural architectures demonstrates that WKLoss achieves an IoU improvement of over 0.1 on the most extreme intensity class while maintaining good calibration, confirming that ordinal supervision significantly outperforms conventional cross-entropy. Nevertheless, prediction of extremely rare events remains constrained by severe sample scarcity.

Technology Category

• Machine Learning: Multi-class/Multi-label Learning & Extreme Classification
• Reasoning under Uncertainty: Stochastic Optimization
• Computer Vision: Learning & Optimization for CV

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Wildfires are highly imbalanced natural hazards in both space and severity, making the prediction of extreme events particularly challenging. In this work, we introduce the first ordinal classification framework for forecasting wildfire severity levels directly aligned with operational decision-making in France. Our study investigates the influence of loss-function design on the ability of neural models to predict rare yet critical high-severity fire occurrences. We compare standard cross-entropy with several ordinal-aware objectives, including the proposed probabilistic TDeGPD loss derived from a truncated discrete exponentiated Generalized Pareto Distribution. Through extensive benchmarking over multiple architectures and real operational data, we show that ordinal supervision substantially improves model performance over conventional approaches. In particular, the Weighted Kappa Loss (WKLoss) achieves the best overall results, with more than +0.1 IoU (Intersection Over Union) gain on the most extreme severity classes while maintaining competitive calibration quality. However, performance remains limited for the rarest events due to their extremely low representation in the dataset. These findings highlight the importance of integrating both severity ordering, data imbalance considerations, and seasonality risk into wildfire forecasting systems. Future work will focus on incorporating seasonal dynamics and uncertainty information into training to further improve the reliability of extreme-event prediction.

Problem

• wildfire prediction
• extreme-value events
• class imbalance
• ordinal classification
• severity forecasting

Innovation

• ordinal classification
• wildfire severity prediction
• imbalanced learning
• loss function design
• extreme-value modeling