Crisp: A Spectral-Based Interaction Strategy for Multivariate Time Series Forecasting
Binwu Wang, Gaoyun Lin, Jiaming Ma, Qihe Huang, Zhengyang Zhou, Xu Wang, Pengkun Wang, Yang Wang
Abstract
Multivariate time series (MTS) forecasting critically depends on modeling inter-variable dependencies, yet existing paradigms face a trade-off: channel-isolation strategies can suffer from information fragmentation in strongly coupled systems, whereas channel-interaction methods often introduce spurious interactions among irrelevant variables. To address this challenge, we propose Coherent Resonance Interaction with Spectral Priors (Crisp). Crisp adopts the principle that effective information exchange should occur only between variables with compatible oscillatory patterns. Concretely, we derive spectral priors in the frequency domain to construct dynamic resonance topologies. With a differentiable, adaptive, and strictly sparse blocking mechanism, Crisp forces attention weights for spectrally inconsistent neighbors to be exactly zero. In addition, we introduce a spectral-gated feature filtering module to refine variable representations using intrinsic spectral characteristics. Extensive experiments demonstrate that Crisp significantly outperforms 20+ baselines. Our code is available at Anonymous GitHub.
BibTeX
@inproceedings{
wang2026crisp,
title={Crisp: A Spectral-Based Interaction Strategy for Multivariate Time Series Forecasting},
author={Binwu Wang and Gaoyun Lin and Jiaming Ma and Qihe Huang and Zhengyang Zhou and Xu Wang and Pengkun Wang and Yang Wang},
booktitle={Forty-third International Conference on Machine Learning},
year={2026},
url={https://openreview.net/forum?id=avYLuILgcr}
}