VisiFold: Long-Term Traffic Forecasting via Temporal Folding Graph and Node Visibility

2026-03-12

AI Summary

Long-term traffic forecasting faces two major challenges: high computational cost and complex spatiotemporal dependencies. Existing methods struggle to model these effectively due to snapshot stacking and fragmented temporal strides. To address this, this work proposes a temporal folding graph structure that integrates multi-step snapshots into a single unified graph. It further introduces a node visibility mechanism combining node-level masking with subgraph sampling, enabling efficient capture of long-range temporal dynamics while substantially reducing computational resource consumption. The approach overcomes the limitations of conventional spatiotemporal graph models that rely on temporal stacking and fully connected spatial structures. Experimental results demonstrate significant performance gains over current baselines in long-term forecasting tasks, maintaining robust accuracy even under a high masking rate of 80%.

Technology Category

• Planning, Routing, and Scheduling: Temporal Planning
• Machine Learning: Graph-based Machine Learning
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG

Abstract

Traffic forecasting is a cornerstone of intelligent transportation systems. While existing research has made significant progress in short-term prediction, long-term forecasting remains a largely uncharted and challenging frontier. Extending the prediction horizon intensifies two critical issues: escalating computational resource consumption and increasingly complex spatial-temporal dependencies. Current approaches, which rely on spatial-temporal graphs and process temporal and spatial dimensions separately, suffer from snapshot-stacking inflation and cross-step fragmentation. To overcome these limitations, we propose \textit{VisiFold}. Our framework introduces a novel temporal folding graph that consolidates a sequence of temporal snapshots into a single graph. Furthermore, we present a node visibility mechanism that incorporates node-level masking and subgraph sampling to overcome the computational bottleneck imposed by large node counts. Extensive experiments show that VisiFold not only drastically reduces resource consumption but also outperforms existing baselines in long-term forecasting tasks. Remarkably, even with a high mask ratio of 80\%, VisiFold maintains its performance advantage. By effectively breaking the resource constraints in both temporal and spatial dimensions, our work paves the way for more realistic long-term traffic forecasting. The code is available at~ https://github.com/PlanckChang/VisiFold.

Problem

• long-term traffic forecasting
• spatial-temporal dependencies
• computational resource consumption
• snapshot-stacking inflation
• cross-step fragmentation

Innovation

• Temporal Folding Graph
• Node Visibility
• Long-Term Traffic Forecasting
• Subgraph Sampling
• Spatial-Temporal Dependency