Tackling the Wildfire Prediction Challenge: An Explainable Artificial Intelligence (XAI) Model Combining Extreme Gradient Boosting (XGBoost) with SHapley Additive exPlanations (SHAP) for Enhanced Interpretability and Accuracy

Forests, Год журнала: 2025, Номер 16(4), С. 689 - 689

Опубликована: Апрель 16, 2025

The intensification of global climate change, combined with increasing human activities, has significantly increased wildfire frequency and severity, posing a major environmental challenge. As an illustration, Guizhou Province in China encountered total 221 wildfires over span 12 days. Despite significant advancements prediction models, challenges related to data imbalance model interpretability persist, undermining their overall reliability. In response these challenges, this study proposes explainable risk (EWXS) leveraging Extreme Gradient Boosting (XGBoost), focus on Province. methodology involved converting raster vector into structured tabular formats, merging, normalizing, encoding them using the Weight Evidence (WOE) technique enhance feature representation. Subsequently, cleaned were balanced establish robust foundation for EWXS model. performance was evaluated comparison established such as CatBoost, range metrics. results indicated that achieved accuracy 99.22%, precision 98.48%, recall 96.82%, F1 score 97.64%, AUC 0.983, thereby demonstrating its strong performance. Moreover, SHAP framework employed interpretability, unveiling key factors influencing risk, including proximity villages, meteorological conditions, air humidity, variations vegetation temperature. This analysis provides valuable support decision-making bodies by offering clear, explanatory insights contributing risk.

Язык: Английский

Identification method of forest fire risk factors and their coupling relationship driven by attribute dependence

International Journal of Disaster Risk Reduction, Год журнала: 2025, Номер unknown, С. 105529 - 105529

Опубликована: Май 1, 2025

Язык: Английский

A new weighted rough set and improved BP neural network method for predicting forest fires

Reliability Engineering & System Safety, Год журнала: 2025, Номер unknown, С. 111206 - 111206

Опубликована: Май 1, 2025

Язык: Английский