Residential segregation assessment based on multi-source data and random forest method: a case study of Nanjing

The assessment of urban residential segregation plays a crucial role in guiding equitable governance and fostering healthy urban development. Existing research evaluating residential segregation from the perspective of wealth disparity often struggles to obtain accurate population distribution data across different socio-economic strata, leading to potential biases in the assessment outcomes. However, residential compounds, as the living spaces for various social groups, can serve as a surrogate for the data representing different socio-economic levels in the evaluation of residential segregation. This paper explores a method that classifies residential compounds based on multi-source data to replace socio-economic strata and assess residential segregation. The experimental results demonstrate that this method can effectively and conveniently evaluate residential segregation, providing a new approach for research in this area.