Traffic Forecasting Comparison Hub

Reference Surfaces

Top Papers

• VisiFold: Long-Term Traffic Forecasting via Temporal Folding Graph and Node Visibility(8.0)

  VisiFold revolutionizes long-term traffic forecasting by integrating temporal folding graphs and node visibility mechanisms to enhance prediction accuracy while reducing computational costs.

• Spatial PDE-aware Selective State-space with Nested Memory for Mobile Traffic Grid Forecasting(6.0)

  NeST-S6 is a novel convolutional selective state-space model for efficient mobile traffic grid forecasting that significantly reduces prediction errors and computational overhead.

• Long-Horizon Traffic Forecasting via Incident-Aware Conformal Spatio-Temporal Transformers(5.0)

  A novel traffic forecasting model that incorporates incident-aware spatio-temporal dynamics for improved accuracy.

• Geographically-aware Transformer-based Traffic Forecasting for Urban Motorway Digital Twins(2.0)

  Develop a Transformer-based traffic forecasting model to enhance real-time motorway traffic management in digital twins.