Flexst improves traffic forecasting across varied data types
A Multi-Resolution Multi-Domain Pre-Training Framework for Universal Traffic Forecasting
Machine Learning
Summary
Traffic data are very different depending on where and when they are collected, which makes predicting traffic tricky. The authors designed FlexST, a flexible system that looks at traffic data in many ways and from different sources to better learn patterns. It uses special parts that focus on short-term changes and long-term trends, and it can tell which parts of the system should handle specific kinds of traffic data. Tests on many datasets showed FlexST can predict traffic better than other methods, even when little data is available.
What this means in practice
- •For urban traffic planners: Create better traffic flow predictions that adapt to various city areas and data collection methods for improved transportation management.
- •For smart city platform developers: Integrate a modular, efficient forecasting model that generalizes across heterogeneous traffic datasets to enhance citywide traffic services.
Authors
Zhouyang Liu, Jindong Han, Hao Wang, Xinyue Liu, Hui Gao, Dongsheng Li, Hao Liu
Abstract
Spatio-temporal traffic data are central to intelligent transportation systems, yet their heterogeneity poses significant challenges for large-scale modeling. Existing pre-trained models often rely on a homogeneous modeling paradigm to handle highly heterogeneous traffic data. This fundamental mismatch not only limits model generalization but also leads to computationally expensive and parameter-inefficient designs. To this end, we propose FlexST, a novel pre-training framework that introduces modularity and adaptivity for traffic modeling. Specifically, we first propose a multi-resolution spatio-temporal diffusion module that captures both short-term fluctuations and long-range trends, effectively reconciling inputs with divergent temporal and spatial resolutions. After that, we construct a domain-adaptive mixture-of-experts that dynamically routes data to specialized sub-networks, enabling selective knowledge transfer while preventing negative interference across diverse domains. Moreover, we devise a unified periodic encoding strategy that injects resolution- and domain-aware inductive biases to harmonize periodic inconsistencies across datasets. Extensive experiments on 23 real-world traffic datasets demonstrate that FlexST significantly outperforms state-of-the-art baselines in zero- and few-shot settings, showcasing superior generalization, adaptability and efficiency. This work offers a new direction for building general-purpose pre-trained models capable of handling the complexity and variability of urban traffic systems.