SDMixer: Sparse Dual-Mixer for Time Series Forecasting
Xiang Ao · 2026
Read SDMixer: Sparse Dual-Mixer for Time Series Forecasting by Xiang Ao (2026) 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-2602-23581.txt.
Description
Xiang Ao offers SDMixer, a sparse dual-mixer method for many-channel time series forecasts. Such forecasts serve transport, energy, and finance, yet data often suffer from multi-scale traits, weak links, and noise, capping current models' forecast scores.
The dual-stream sparse Mixer extracts global trends and local dynamic traits from series in both rate and time domains. A sparing filter drops invalid data, lifting scores of cross-variable link building. Test results show the method gets leading scores on multiple real-world data sets, proving its worth and breadth. Code is shared openly.
Catalog record
| CREATOR | Xiang Ao |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 25 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-23581