Decorrelating the Future: Joint Frequency Domain Learning for Spatio-temporal Forecasting
This addresses forecasting accuracy for spatio-temporal data, offering an incremental improvement by extending frequency-domain methods to joint spatio-temporal interactions.
The paper tackled the problem of capturing complex spatio-temporal dependencies in graph-structured signals for forecasting, proposing a frequency-enhanced training objective that improved state-of-the-art baselines across six real-world datasets.
Standard direct forecasting models typically rely on point-wise objectives such as Mean Squared Error, which fail to capture the complex spatio-temporal dependencies inherent in graph-structured signals. While recent frequency-domain approaches such as FreDF mitigate temporal autocorrelation, they often overlook spatial and cross spatio-temporal interactions. To address this limitation, we propose FreST Loss, a frequency-enhanced spatio-temporal training objective that extends supervision to the joint spatio-temporal spectrum. By leveraging the Joint Fourier Transform (JFT), FreST Loss aligns model predictions with ground truth in a unified spectral domain, effectively decorrelating complex dependencies across both space and time. Theoretical analysis shows that this formulation reduces estimation bias associated with time-domain training objectives. Extensive experiments on six real-world datasets demonstrate that FreST Loss is model-agnostic and consistently improves state-of-the-art baselines by better capturing holistic spatio-temporal dynamics.