Fusion Matrix Prompt Enhanced Self-Attention Spatial-Temporal Interactive Traffic Forecasting Framework

TL;DR

This paper introduces FMPESTF, a novel traffic forecasting framework that combines spatial-temporal fusion matrices and attention mechanisms to improve accuracy by capturing complex traffic correlations and dynamics.

Contribution

The paper proposes a new spatial-temporal interactive framework with fusion matrices and hierarchical learning, addressing limitations of previous models in traffic correlation and time importance.

Findings

Outperforms baseline models on six real-world datasets

Demonstrates high accuracy and efficiency in traffic forecasting

Effectively captures spatial-temporal heterogeneity and node relationships

Abstract

Recently, spatial-temporal forecasting technology has been rapidly developed due to the increasing demand for traffic management and travel planning. However, existing traffic forecasting models still face the following limitations. On one hand, most previous studies either focus too much on real-world geographic information, neglecting the potential traffic correlation between different regions, or overlook geographical position and only model the traffic flow relationship. On the other hand, the importance of different time slices is ignored in time modeling. Therefore, we propose a Fusion Matrix Prompt Enhanced Self-Attention Spatial-Temporal Interactive Traffic Forecasting Framework (FMPESTF), which is composed of spatial and temporal modules for down-sampling traffic data. The network is designed to establish a traffic fusion matrix considering spatial-temporal heterogeneity as a query to reconstruct a data-driven dynamic traffic data structure, which accurately reveal the flow relationship of nodes in the traffic network. In addition, we introduce attention mechanism in time modeling, and design hierarchical spatial-temporal interactive learning to help the model adapt to various traffic scenarios. Through extensive experimental on six real-world traffic datasets, our method is significantly superior to other baseline models, demonstrating its efficiency and accuracy in dealing with traffic forecasting problems.