On the Prediction of a Multidimensional Vulnerability Index for Climate Resilience

As regions across the globe become more susceptible to natural disasters, economic shocks, and climate change, the accurate prediction of their vulnerability is becoming more essential for resilience planning. This study proposes an innovative application of machine learning modeling techniques to forecast the Multidimensional Vulnerability Index scores for climate-vulnerable countries, especially those located within the Caribbean. While previous studies have utilized traditional methods for calculating vulnerability indices or done so on smaller scales, our approach leverages machine learning to demonstrate its predictive accuracy and explore the relationships between various predictors influencing vulnerability. This paper presents the first comprehensive attempt to apply machine learning to vulnerability prediction on the international scale, offering a data-driven approach to guide policy decisions in disaster risk management and climate resilience planning.