RDIT: Residual-based Diffusion Implicit Models for Probabilistic Time Series Forecasting
This work addresses uncertainty-aware predictions for decision-making in domains like finance or healthcare, though it appears incremental as it builds on existing point estimators and diffusion methods.
The paper tackles probabilistic time series forecasting by proposing RDIT, a framework that combines point estimation with residual-based conditional diffusion and a bidirectional Mamba network, achieving state-of-the-art performance with lower CRPS, rapid inference, and improved coverage across eight multivariate datasets.
Probabilistic Time Series Forecasting (PTSF) plays a critical role in domains requiring accurate and uncertainty-aware predictions for decision-making. However, existing methods offer suboptimal distribution modeling and suffer from a mismatch between training and evaluation metrics. Surprisingly, we found that augmenting a strong point estimator with a zero-mean Gaussian, whose standard deviation matches its training error, can yield state-of-the-art performance in PTSF. In this work, we propose RDIT, a plug-and-play framework that combines point estimation and residual-based conditional diffusion with a bidirectional Mamba network. We theoretically prove that the Continuous Ranked Probability Score (CRPS) can be minimized by adjusting to an optimal standard deviation and then derive algorithms to achieve distribution matching. Evaluations on eight multivariate datasets across varied forecasting horizons demonstrate that RDIT achieves lower CRPS, rapid inference, and improved coverage compared to strong baselines.