Ister: Linear Transformer for Efficient Multivariate Time Series Forecasting

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Cao, Fanpu, Yang, Shu, Chen, Zhengjian, Liu, Ye, Cui, Laizhong
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914546707333120
author Cao, Fanpu
Yang, Shu
Chen, Zhengjian
Liu, Ye
Cui, Laizhong
author_facet Cao, Fanpu
Yang, Shu
Chen, Zhengjian
Liu, Ye
Cui, Laizhong
contents Transformer-based models have achieved remarkable success in multivariate time series forecasting (MTSF) by capturing long-range dependencies. However, their widespread adoption is hindered by the quadratic computational complexity of self-attention, which limits scalability on high-dimensional sequences. To address this challenge, we propose the Inverted Seasonal-Trend Decomposition Transformer (Ister), a novel architecture that enhances both predictive accuracy and computational efficiency. Central to Ister is Dot-attention, a linear-complexity attention mechanism that replaces conventional multi-head self-attention with element-wise dot-product operations to model inter-series dependencies. Furthermore, we introduce an inverted seasonal-trend decomposition strategy that isolates periodic components, enabling the model to focus learning on periodic patterns, thereby improving the performance of channel alignment. Extensive experiments across several real-world benchmarks demonstrate that Ister consistently achieves state-of-the-art performance. Code is available at https://github.com/macovaseas/Ister.
format Preprint
id arxiv_https___arxiv_org_abs_2412_18798
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Ister: Linear Transformer for Efficient Multivariate Time Series Forecasting
Cao, Fanpu
Yang, Shu
Chen, Zhengjian
Liu, Ye
Cui, Laizhong
Machine Learning
Artificial Intelligence
Transformer-based models have achieved remarkable success in multivariate time series forecasting (MTSF) by capturing long-range dependencies. However, their widespread adoption is hindered by the quadratic computational complexity of self-attention, which limits scalability on high-dimensional sequences. To address this challenge, we propose the Inverted Seasonal-Trend Decomposition Transformer (Ister), a novel architecture that enhances both predictive accuracy and computational efficiency. Central to Ister is Dot-attention, a linear-complexity attention mechanism that replaces conventional multi-head self-attention with element-wise dot-product operations to model inter-series dependencies. Furthermore, we introduce an inverted seasonal-trend decomposition strategy that isolates periodic components, enabling the model to focus learning on periodic patterns, thereby improving the performance of channel alignment. Extensive experiments across several real-world benchmarks demonstrate that Ister consistently achieves state-of-the-art performance. Code is available at https://github.com/macovaseas/Ister.
title Ister: Linear Transformer for Efficient Multivariate Time Series Forecasting
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2412.18798