A New Approach to the Social Vulnerability Indices: Decision Tree-Based Vulnerability Classification Model

The Social Vulnerability Index (SVI), a composite score identifying populations at risk from disasters, is often used to predict vulnerability and plan for community-based disaster prevention and emergency response. Our study introduces a decision tree based approach to developing an SVI that captures the heterogeneity of both vulnerable populations and disasters and we demonstrate the importance of incorporating a disaster loss classification into estimating social vulnerability to increase the predictive performance of the model. Findings suggest that the SVI based on the decision tree approach dramatically increased the accuracy of predicting high vulnerability areas.