This research explores the potential of urban networks to model, analyse, and visualise the structure and flow of complex urban systems. It makes three primary contributions: 1) re-evaluating the theoretical foundations of urban networks through the integration of graph machine learning, urban complexity, and network sciences; 2) harnessing real-world datasets, including volunteered geographic and remote sensing open data, for city scale urban sensing and multi-dimensional representation learning on urban graphs; and 3) developing practical deep learning based workflows for evidence-based planning and urban sustainability.
Through a data-driven perspective, the study aims to advance an understanding of the structure and dynamics of urban networks at an unprecedented scale and resolution.
