DPANet: Dual Pyramid Attention Network for Multivariate Time Series Forecasting

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Li, Qianyang, Zhang, Xingjun, Wang, Shaoxun, Wei, Jia
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912594113069056
author Li, Qianyang
Zhang, Xingjun
Wang, Shaoxun
Wei, Jia
author_facet Li, Qianyang
Zhang, Xingjun
Wang, Shaoxun
Wei, Jia
contents Long-term time series forecasting (LTSF) is hampered by the challenge of modeling complex dependencies that span multiple temporal scales and frequency resolutions. Existing methods, including Transformer and MLP-based models, often struggle to capture these intertwined characteristics in a unified and structured manner. We propose the Dual Pyramid Attention Network (DPANet), a novel architecture that explicitly decouples and concurrently models temporal multi-scale dynamics and spectral multi-resolution periodicities. DPANet constructs two parallel pyramids: a Temporal Pyramid built on progressive downsampling, and a Frequency Pyramid built on band-pass filtering. The core of our model is the Cross-Pyramid Fusion Block, which facilitates deep, interactive information exchange between corresponding pyramid levels via cross-attention. This fusion proceeds in a coarse-to-fine hierarchy, enabling global context to guide local representation learning. Extensive experiments on public benchmarks show that DPANet achieves state-of-the-art performance, significantly outperforming prior models. Code is available at https://github.com/hit636/DPANet.
format Preprint
id arxiv_https___arxiv_org_abs_2509_14868
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle DPANet: Dual Pyramid Attention Network for Multivariate Time Series Forecasting
Li, Qianyang
Zhang, Xingjun
Wang, Shaoxun
Wei, Jia
Machine Learning
Artificial Intelligence
Long-term time series forecasting (LTSF) is hampered by the challenge of modeling complex dependencies that span multiple temporal scales and frequency resolutions. Existing methods, including Transformer and MLP-based models, often struggle to capture these intertwined characteristics in a unified and structured manner. We propose the Dual Pyramid Attention Network (DPANet), a novel architecture that explicitly decouples and concurrently models temporal multi-scale dynamics and spectral multi-resolution periodicities. DPANet constructs two parallel pyramids: a Temporal Pyramid built on progressive downsampling, and a Frequency Pyramid built on band-pass filtering. The core of our model is the Cross-Pyramid Fusion Block, which facilitates deep, interactive information exchange between corresponding pyramid levels via cross-attention. This fusion proceeds in a coarse-to-fine hierarchy, enabling global context to guide local representation learning. Extensive experiments on public benchmarks show that DPANet achieves state-of-the-art performance, significantly outperforming prior models. Code is available at https://github.com/hit636/DPANet.
title DPANet: Dual Pyramid Attention Network for Multivariate Time Series Forecasting
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2509.14868