Holistic and Scalable Approaches in Transportation and Mobility Prediction

The recent developments in the research area of transportation and mobility prediction have seen significant advancements, particularly in the integration of deep learning models with spatial-temporal data. A notable trend is the shift towards more holistic and scalable models that can handle diverse and complex urban environments. These models often leverage novel architectures such as transformers and graph neural networks to capture intricate patterns and dependencies in data, which is crucial for accurate predictions. Additionally, there is a growing emphasis on incorporating long-term historical data and dynamic influences to enhance the robustness and reliability of prediction models. The field is also witnessing innovations in data fusion techniques, where multiple types of data, including spatial, temporal, and contextual factors, are combined to improve prediction accuracy. Furthermore, the use of foundation models and meta-learning approaches is gaining traction, enabling models to generalize better across different tasks and datasets, thereby reducing the need for extensive retraining and improving adaptability. These advancements are paving the way for more efficient and intelligent transportation systems, with potential applications ranging from traffic management to public transit optimization and electric vehicle infrastructure planning.

Noteworthy papers include 'SatQA: A New Model for Satellite Streaming Video QoE Prediction Using Network Parameters' for its innovative approach to predicting video quality without pixel data, and 'ST-MoE-BERT: A Spatial-Temporal Mixture-of-Experts Framework for Long-Term Cross-City Mobility Prediction' for its robust cross-city mobility prediction capabilities.

Sources

Satellite Streaming Video QoE Prediction: A Real-World Subjective Database and Network-Level Prediction Models

ST-MoE-BERT: A Spatial-Temporal Mixture-of-Experts Framework for Long-Term Cross-City Mobility Prediction

Leveraging Intra-Period and Inter-Period Features for Enhanced Passenger Flow Prediction of Subway Stations

Incorporating Long-term Data in Training Short-term Traffic Prediction Model

ArrivalNet: Predicting City-wide Bus/Tram Arrival Time with Two-dimensional Temporal Variation Modeling

DST-TransitNet: A Dynamic Spatio-Temporal Deep Learning Model for Scalable and Efficient Network-Wide Prediction of Station-Level Transit Ridership

Multi-class within-day dynamic traffic equilibrium with strategic travel time information

FoMo: A Foundation Model for Mobile Traffic Forecasting with Diffusion Model

SSMT: Few-Shot Traffic Forecasting with Single Source Meta-Transfer

Modelling Concurrent RTP Flows for End-to-end Predictions of QoS in Real Time Communications

Congestion Forecast for Trains with Railroad-Graph-based Semi-Supervised Learning using Sparse Passenger Reports

Anomaly Resilient Temporal QoS Prediction using Hypergraph Convoluted Transformer Network

TranSPORTmer: A Holistic Approach to Trajectory Understanding in Multi-Agent Sports

Attention-based Citywide Electric Vehicle Charging Demand Prediction Approach Considering Urban Region and Dynamic Influences

Built with on top of