A Review of Graph-Based Traffic Flow Prediction: Methods and Challenges

Driven by unprecedented urban sprawl and surging transport needs, accurate traffic prediction now constitutes a fundamental pillar of advanced transportation management systems, with profound implications for metropolitan traffic policy-making, system optimization, and real-time regulation. Currently, the problems existing in traffic flow prediction mainly fall into two categories: information collection and prediction. The intricate topology of urban transportation systems coupled with volatile traffic demand fluctuations results in compromised data acquisition accuracy. Moreover, during the evolution of traffic forecasting techniques methods, there are also issues of imprecise predictions. To systematically comprehend the evolution of traffic forecasting methodologies, critically assess the merits and limitations of prevailing techniques, and enhance subsequent predictive frameworks, this study presents a comprehensive historical analysis of traffic flow prediction development.