Regional railway development disparities in China based on deep embedded clustering and RBF network

Due to uneven regional development in China, significant disparities have emerged in railway development across different regions. This not only constrains the overall efficiency of the railway network but also hinders the advancement of regional coordination strategies. This paper proposes an integrated analytical framework combining deep embedding clustering with radial basis function (RBF) neural networks, offering a novel approach to deciphering the underlying logic of regional differences in railway development. The study first employs deep embedding clustering to integrate multidimensional indicators to cluster China's 31 provinces into four distinct economic regions, laying the foundation for subsequent differentiated analysis. Subsequently, RBF neural network models are constructed for each region to fit railway development trajectories and predict future trends through data training. Results reveal three distinct railway development patterns across the four regions: “positive stabilization,” “limited promotion,” and “negative suppression.” These findings clearly demonstrate the nonlinear relationship between macro-regional economic characteristics and railway development. The data analysis system and conclusions of this study not only offer an innovative data-driven perspective for understanding the intrinsic driving mechanisms of regional transportation infrastructure but also provide scientific support for formulating differentiated and precise regional transportation policy planning at both national and local levels.