A Traffic Prediction Model Based on Multi Stream Feature Fusion

Traffic is constantly changing, traffic supposition is extremely crucial in urban planning congestion control as well as management systems. Long-term traffic predictions help with relevant factors and smart city current traffic reduction. Traffic flow is an important topic because it transports regional factors and patterns regional spatial structure. Traffic prediction is gaining popularity due to its widespread real-world applications in traffic management, urban computing, public safety, and other fields. For intelligent transportation systems, timely and accurate traffic flow prediction is critical. A MSFF-Multi-Stream-Feature-Fusion method is proposed in this paper to remove and incorporate rich features after traffic data.