A knowledge-informed dynamic correlation modeling framework for lane-level traffic flow prediction

Computer Science

• Modeling Framework 100%
• Traffic Flow Prediction 100%
• Large Language Model 100%
• Traffic Prediction 60%
• Computational Cost 40%
• Spatial Correlation 40%
• Attention (Machine Learning) 20%
• Prediction Accuracy 20%
• Superior Performance 20%
• Autonomous Vehicles 20%
• Deep Learning Model 20%
• Dynamic Traffic 20%
• Small Sample Size 20%
• Resource Demand 20%
• Spatial Interaction 20%
• Prediction Function 20%
• Training Sample 20%
• Traffic Management 20%