Cross Space and Time: A Spatio-Temporal Unitized Model for Traffic Flow Forecasting

TL;DR

This paper introduces STUM, a unified spatio-temporal model that captures complex interactions in traffic flow forecasting, improving accuracy and efficiency by integrating novel modules and techniques.

Contribution

The paper presents the Adaptive Spatio-temporal Unitized Cell (ASTUC) and a modular framework that enhances spatio-temporal traffic prediction by capturing heterogeneity and dependencies more effectively.

Findings

STUM outperforms existing models on multiple datasets.

The framework achieves high accuracy with low computational cost.

Hyperparameter tuning and pre-training improve performance further.

Abstract

Predicting spatio-temporal traffic flow presents significant challenges due to complex interactions between spatial and temporal factors. Existing approaches often address these dimensions in isolation, neglecting their critical interdependencies. In this paper, we introduce the Spatio-Temporal Unitized Model (STUM), a unified framework designed to capture both spatial and temporal dependencies while addressing spatio-temporal heterogeneity through techniques such as distribution alignment and feature fusion. It also ensures both predictive accuracy and computational efficiency. Central to STUM is the Adaptive Spatio-temporal Unitized Cell (ASTUC), which utilizes low-rank matrices to seamlessly store, update, and interact with space, time, as well as their correlations. Our framework is also modular, allowing it to integrate with various spatio-temporal graph neural networks through components such as backbone models, feature extractors, residual fusion blocks, and predictive modules to collectively enhance forecasting outcomes. Experimental results across multiple real-world datasets demonstrate that STUM consistently improves prediction performance with minimal computational cost. These findings are further supported by hyperparameter optimization, pre-training analysis, and result visualization. We provide our source code for reproducibility at https://anonymous.4open.science/r/STUM-E4F0.