<ul data-eligibleForWebStory="true"><li>Multivariate long-term time series forecasting faces challenges in capturing temporal dependencies and spatial correlations simultaneously.</li><li>Current approaches like Transformers do not address time series properties like periodicity effectively.</li><li>FNF is introduced as a dedicated backbone and DBD as the architecture for spatio-temporal modeling.</li><li>FNF unifies local time-domain and global frequency-domain information processing within a single backbone, extending to spatial modeling.</li><li>DBD offers superior gradient flow and representation capacity.</li><li>Empirical evaluation across 11 public benchmark datasets in various domains demonstrates state-of-the-art performance.</li><li>The approach achieves results without auxiliary techniques, indicating the potential for improved time series modeling in scientific and industrial applications.</li></ul>

Multivariate Long-term Time Series Forecasting with Fourier Neural Filter

Discover more