PREDICTION OF MISERY INDEX LEVERAGING CLIMATE AND ECONOMIC DATA USING RANDOM FOREST

The Misery Index, a traditional measure of economic distress based income threshold, which provides a limited understanding of socioeconomic realities, particularly in vulnerable regions such as SubSaharan Africa. This region faces unique challenges, as climate change exacerbates economic instability through its impact on agriculture, health, and livelihoods. Despite these significant effects, conventional Misery Index frameworks fail to account for climatic factors, leaving policymakers without a holistic tool to evaluate economic and environmental challenges. To address this gap, this study incorporates climate variables into the Misery Index using a Random Forest model, a machine learning algorithm known for its robustness and ability to handle complex, high-dimensional datasets. Key climatic variables, including temperature, wind speed, humidity, cloud cover, precipitation, ocean heat content, CO₂ levels, ocean pH, sea level, glacier mass balance, Arctic sea ice extent, and Antarctic sea ice extent, were analyzed using Principal Component Analysis (PCA) to derive a representative climate index. The first principal component (PC1), which explained 15.32% of the total variance, was combined with traditional economic variables to enhance the Misery Index's explanatory power. The Random Forest model demonstrated strong predictive performance, achieving a Mean Squared Error (MSE) of 0.5684 and an R-squared value of 0.9953. The benefits of integrating climatic and economic parameters, helps create a framework for understanding the Misery Index in the context of SubSaharan Africa. The model equips policymakers with a dynamic tool to assess the multifaceted impacts of climate change and economic distress, fostering more informed decision-making.