T-Graphormer: Using Transformers for Spatiotemporal Forecasting
Hao Yuan Bai, Xue Liu · 2025
Read T-Graphormer: Using Transformers for Spatiotemporal Forecasting by Hao Yuan Bai, Xue Liu (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-13274.txt.
Description
Hao Yuan Bai and Xue Liu present a forecasting paper, published in 2025, on data that varies across both places and time. Traffic flow is a prime example, where each sensor's future depends on its neighbors and on past patterns at once. Older methods usually model the space links and the time links in separate steps, which can miss how they interact.
Their model, T-Graphormer, adapts a graph aware transformer so every point in a sequence can attend to every other point, with time stamps woven into the input. Nothing about the layout must be fixed in advance. On real traffic benchmark data the method cut standard error measures by a clear margin over leading rivals.
Catalog record
| CREATOR | Hao Yuan Bai, Xue Liu |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 47 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-13274