PMDformer: Patch-Mean Decoupling Information Transformer for Long-term Forecasting
Ao Hu, Liangjian Wen, Jiang Duan, Yong Dai, HE YAN, Dongkai Wang, Jun Wang, Yukun Zhang, Ruoxi Jiang, Zenglin Xu, [0.5em] Southwestern University of Finance, Economics, [0.3em] Chengdu Everimaging Science, Technology Co, Ltd · 2026
Read PMDformer: Patch-Mean Decoupling Information Transformer for Long-term Forecasting by Ao Hu, Liangjian Wen, Jiang Duan, Yong Dai, HE YAN, Dongkai Wang, Jun Wang, Yukun Zhang, Ruoxi Jiang, Zenglin Xu, [0.5em] Southwestern University of Finance, Economics, [0.3em] Chengdu Everimaging Science, Technology Co, Ltd (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-2606-26549.txt.
Description
Ao Hu and nine colleagues present PMDformer, a transformer model for long-term time series forecasting. Forecasting far ahead matters for energy, finance, and traffic planning. Transformer models often chop series into patches, short segments treated as units. Comparing patch shapes across variables is hard because scales differ. A tall patch and a short patch can share the same shape.
Their patch-mean decoupling subtracts each patch's average before comparison. Shape information then stands apart from level, so attention compares true shapes. Two attention modules complete the design. One restores the removed trend when producing output. The other limits cross-variable attention to the most relevant recent segments, avoiding stale correlations. Wide experiments show PMDformer beating leading rivals on stability and accuracy, the authors report.
Catalog record
| CREATOR | Ao Hu, Liangjian Wen, Jiang Duan, Yong Dai, HE YAN, Dongkai Wang, Jun Wang, Yukun Zhang, Ruoxi Jiang, Zenglin Xu, [0.5em] Southwestern University of Finance, Economics, [0.3em] Chengdu Everimaging Science, Technology Co, Ltd |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 40 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.26549