Saved in:
Bibliographic Details
Main Authors: Yang, Qingyuan, Deng, Shizhuo, Chen, Dongyue, Teng, Da, Gan, Zehua
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2512.07539
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909950355177472
author Yang, Qingyuan
Deng, Shizhuo
Chen, Dongyue
Teng, Da
Gan, Zehua
author_facet Yang, Qingyuan
Deng, Shizhuo
Chen, Dongyue
Teng, Da
Gan, Zehua
contents Traditional Transformers face a major bottleneck in long-sequence time series forecasting due to their quadratic complexity $(\mathcal{O}(T^2))$ and their limited ability to effectively exploit frequency-domain information. Inspired by RWKV's $\mathcal{O}(T)$ linear attention and frequency-domain modeling, we propose FRWKV, a frequency-domain linear-attention framework that overcomes these limitations. Our model integrates linear attention mechanisms with frequency-domain analysis, achieving $\mathcal{O}(T)$ computational complexity in the attention path while exploiting spectral information to enhance temporal feature representations for scalable long-sequence modeling. Across eight real-world datasets, FRWKV achieves a first-place average rank. Our ablation studies confirm the critical roles of both the linear attention and frequency-encoder components. This work demonstrates the powerful synergy between linear attention and frequency analysis, establishing a new paradigm for scalable time series modeling. Code is available at this repository: https://github.com/yangqingyuan-byte/FRWKV.
format Preprint
id arxiv_https___arxiv_org_abs_2512_07539
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle FRWKV:Frequency-Domain Linear Attention for Long-Term Time Series Forecasting
Yang, Qingyuan
Deng, Shizhuo
Chen, Dongyue
Teng, Da
Gan, Zehua
Machine Learning
Traditional Transformers face a major bottleneck in long-sequence time series forecasting due to their quadratic complexity $(\mathcal{O}(T^2))$ and their limited ability to effectively exploit frequency-domain information. Inspired by RWKV's $\mathcal{O}(T)$ linear attention and frequency-domain modeling, we propose FRWKV, a frequency-domain linear-attention framework that overcomes these limitations. Our model integrates linear attention mechanisms with frequency-domain analysis, achieving $\mathcal{O}(T)$ computational complexity in the attention path while exploiting spectral information to enhance temporal feature representations for scalable long-sequence modeling. Across eight real-world datasets, FRWKV achieves a first-place average rank. Our ablation studies confirm the critical roles of both the linear attention and frequency-encoder components. This work demonstrates the powerful synergy between linear attention and frequency analysis, establishing a new paradigm for scalable time series modeling. Code is available at this repository: https://github.com/yangqingyuan-byte/FRWKV.
title FRWKV:Frequency-Domain Linear Attention for Long-Term Time Series Forecasting
topic Machine Learning
url https://arxiv.org/abs/2512.07539