Spatial distribution patterns of urban economic activities: a graph convolutional network approach

Abstract

With the acceleration of urbanization, the spatial distribution patterns of urban economic activities are becoming increasingly complex. How to scientifically and effectively analyze and optimize these distribution patterns has become one of the key issues in urban planning and economic development. Based on spatial data analysis and Graph Convolutional Networks (GCNs), this paper investigates the distribution patterns and influencing factors of urban economic activities. By collecting multi-dimensional spatial data and employing GIS, spatial statistics, and GCN-based modeling, we reveal the spatial aggregation patterns and diffusion mechanisms of different economic activities in urban areas. The results show that urban economic activities show obvious spatial agglomeration effect, especially in the core area and near transportation hubs, where economic activities are more concentrated. Further analysis shows that factors such as land use type, traffic network density and population density have significant influence on the spatial distribution of economic activities. In the specific data analysis, 10 city sample data were used, and was quantitatively evaluated by spatial autocorrelation analysis and regression model. The results show that the agglomeration effect of economic activities is closely related to the scale of cities and the perfection of transportation facilities, and big cities are more likely to form high-density economic areas than small cities. The research in this paper provides important theoretical basis and data support for urban planners and policy makers, which is helpful to optimize the spatial layout activities and improve the efficiency of resource allocation. Finally, the article puts forward several policy suggestions to promote the rationalization of urban economic space, promote the sustainable development of cities.