Guardado en:
Detalles Bibliográficos
Autores principales: Xu, Jiyuan, Zhang, Wenyu, Jing, Xin, Chen, Shuai, Zhang, Shuai, Nie, Jiahao
Formato: Preprint
Publicado: 2026
Materias:
Acceso en línea:https://arxiv.org/abs/2601.20318
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866914355253084160
author Xu, Jiyuan
Zhang, Wenyu
Jing, Xin
Chen, Shuai
Zhang, Shuai
Nie, Jiahao
author_facet Xu, Jiyuan
Zhang, Wenyu
Jing, Xin
Chen, Shuai
Zhang, Shuai
Nie, Jiahao
contents Current methods for multivariate time series forecasting can be classified into channel-dependent and channel-independent models. Channel-dependent models learn cross-channel features but often overfit the channel ordering, which hampers adaptation when channels are added or reordered. Channel-independent models treat each channel in isolation to increase flexibility, yet this neglects inter-channel dependencies and limits performance. To address these limitations, we propose \textbf{CPiRi}, a \textbf{channel permutation invariant (CPI)} framework that infers cross-channel structure from data rather than memorizing a fixed ordering, enabling deployment in settings with structural and distributional co-drift without retraining. CPiRi couples \textbf{spatio-temporal decoupling architecture} with \textbf{permutation-invariant regularization training strategy}: a frozen pretrained temporal encoder extracts high-quality temporal features, a lightweight spatial module learns content-driven inter-channel relations, while a channel shuffling strategy enforces CPI during training. We further \textbf{ground CPiRi in theory} by analyzing permutation equivariance in multivariate time series forecasting. Experiments on multiple benchmarks show state-of-the-art results. CPiRi remains stable when channel orders are shuffled and exhibits strong \textbf{inductive generalization} to unseen channels even when trained on \textbf{only half} of the channels, while maintaining \textbf{practical efficiency} on large-scale datasets. The source code is released at https://github.com/JasonStraka/CPiRi.
format Preprint
id arxiv_https___arxiv_org_abs_2601_20318
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle CPiRi: Channel Permutation-Invariant Relational Interaction for Multivariate Time Series Forecasting
Xu, Jiyuan
Zhang, Wenyu
Jing, Xin
Chen, Shuai
Zhang, Shuai
Nie, Jiahao
Computer Vision and Pattern Recognition
Current methods for multivariate time series forecasting can be classified into channel-dependent and channel-independent models. Channel-dependent models learn cross-channel features but often overfit the channel ordering, which hampers adaptation when channels are added or reordered. Channel-independent models treat each channel in isolation to increase flexibility, yet this neglects inter-channel dependencies and limits performance. To address these limitations, we propose \textbf{CPiRi}, a \textbf{channel permutation invariant (CPI)} framework that infers cross-channel structure from data rather than memorizing a fixed ordering, enabling deployment in settings with structural and distributional co-drift without retraining. CPiRi couples \textbf{spatio-temporal decoupling architecture} with \textbf{permutation-invariant regularization training strategy}: a frozen pretrained temporal encoder extracts high-quality temporal features, a lightweight spatial module learns content-driven inter-channel relations, while a channel shuffling strategy enforces CPI during training. We further \textbf{ground CPiRi in theory} by analyzing permutation equivariance in multivariate time series forecasting. Experiments on multiple benchmarks show state-of-the-art results. CPiRi remains stable when channel orders are shuffled and exhibits strong \textbf{inductive generalization} to unseen channels even when trained on \textbf{only half} of the channels, while maintaining \textbf{practical efficiency} on large-scale datasets. The source code is released at https://github.com/JasonStraka/CPiRi.
title CPiRi: Channel Permutation-Invariant Relational Interaction for Multivariate Time Series Forecasting
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2601.20318