U-Former ODE: Fast Probabilistic Forecasting of Irregular Time Series
This addresses a scalability bottleneck in probabilistic forecasting for domains like healthcare and finance, offering significant speed improvements while maintaining accuracy.
The paper tackled the challenge of slow, sequential computation in probabilistic forecasting of irregular time series by introducing UFO, a novel architecture that integrates U-Nets, Transformers, and Neural CDEs. It achieved up to 15x faster inference and outperformed ten state-of-the-art baselines in predictive accuracy across five benchmarks.
Probabilistic forecasting of irregularly sampled time series is crucial in domains such as healthcare and finance, yet it remains a formidable challenge. Existing Neural Controlled Differential Equation (Neural CDE) approaches, while effective at modelling continuous dynamics, suffer from slow, inherently sequential computation, which restricts scalability and limits access to global context. We introduce UFO (U-Former ODE), a novel architecture that seamlessly integrates the parallelizable, multiscale feature extraction of U-Nets, the powerful global modelling of Transformers, and the continuous-time dynamics of Neural CDEs. By constructing a fully causal, parallelizable model, UFO achieves a global receptive field while retaining strong sensitivity to local temporal dynamics. Extensive experiments on five standard benchmarks -- covering both regularly and irregularly sampled time series -- demonstrate that UFO consistently outperforms ten state-of-the-art neural baselines in predictive accuracy. Moreover, UFO delivers up to 15$\times$ faster inference compared to conventional Neural CDEs, with consistently strong performance on long and highly multivariate sequences.