SIMBAD: Spatio-Temporal Traffic Forecasting Robust to Aperiodicity

arXiv:2611.09015 · cs.AI · Submitted 2026-11-06 · Read on arXiv

cs.AI

Submitted: 2026-11-06

Updated: 2026-11-06

Comments: Total of 30 pages, 9 figures (4 in main, 5 in appendix), accepted to NeurIPS(Neural Information Processing Systems) 2026

Code: https://github.com/theghks3/SIMBAD

License: http://creativecommons.org/licenses/by/4.0/

The gist: Accurate traffic prediction is essential for urban planning but remains challenging due to the irregular patterns caused by unexpected events.

Abstract

Accurate traffic prediction is essential for urban planning but remains challenging due to the irregular patterns caused by unexpected events. While Spatio-Temporal Graph Neural Networks (STGNNs) effectively model periodic data, they often struggle with these aperiodic fluctuations. To address this, we propose SIMBAD, which improves generalization irregular patterns via two key mechanisms: 1) adaptive regulation of periodic signals to prioritize recent signals, and 2) dynamic spatial modeling that adjusts influence between connected nodes based on contextual relations. Built upon simple similarity metrics to adapt to aperiodic patterns, SIMBAD outperforms existing methods on real-world benchmark datasets, demonstrating superior performance in predicting unseen and irregular traffic events.

Sources

Related papers