FOREcaST: Improving Extreme Weather Forecasts with Deep Neural Decision Forest for Climate Change Adaptation

Climate change poses significant challenges for society, particularly in mitigating the impacts of extreme weather events. Accurate and timely forecasts of extreme weather phenomena are crucial for effective adaptation strategies to cope with disasters and minimize serious consequences. This paper presents a hybrid approach, FOREcaST, for enhanced extreme weather forecasting, which leverages the power of deep neural networks and decision forests to improve the prediction accuracy of extreme weather events. Specifically, we proposed using the Deep Neural Decision Forest for the regression problem and approaches for customizing and ensembling DNDF Regression. The proposed framework enables a better understanding of the complex relationships within weather data and enhances the prediction capabilities for various extreme weather events. The experiments conducted on real-world datasets demonstrate the performance of FOREcaST over existing forecasting methods, highlighting its potential to support decision-making processes and improve climate change adaptation strategies.