Machine learning framework for multidimensional assessment of urban quality of life.

This study uses statistical and machine learning techniques to categorize and rank 99 high development cities according to multidimensional Quality of Life factors (QoL). We categorize cities into three unique clusters using hierarchical Ward.D2 clustering and Principal Component Analysis for dimension reduction. We discover clusters that bring together economically developed cities with strong social safety nets, high-income cities lacking in public amenities, rising cities of the future located in peripheral regions, and major population centers in the developing world, finding considerable structural similarities that lead to comparable outcomes across culturally and geographically diverse cities. Components of QoL are grouped into three Principal Components that reveal dynamic interactions between different variables influencing economic and experimental aspects of QoL. Clusters are assessed relative to one another using the three PCs, and crucial factors for classifying cities into different clusters are identified by a Decision tree to allow for tailored policy recommendations for cities of different clusters. Our findings give policymakers a framework to prioritize holistic urban development over GDP-centric models, highlighting the importance of striking a balance between objective standards of human development and subjective, experiential indicators of individual wellbeing.