Search papers, labs, and topics across Lattice.
This paper introduces VLBM (Variational Latent Basis Model), a novel framework designed to enhance out-of-distribution (OOD) robustness in multivariate time series forecasting by separating stable dynamics from OOD-induced deviations. By employing a shared latent basis for stable in-distribution dynamics and aligning future-aware posteriors with future-blind priors, VLBM effectively mitigates the overwhelming influence of frequent in-distribution patterns during training. The model demonstrates significant improvements in OOD robustness and in-distribution accuracy across 12 benchmark tasks, achieving average MAE and MSE gains of 15.08% and 7.74% over the strongest baseline.
VLBM achieves state-of-the-art OOD robustness while maintaining high in-distribution accuracy, outperforming existing methods by a notable margin.
Out of distribution (OOD) events in multivariate time series forecasting are rare but often dominate real world risk, making average case forecasting insufficient for reliable deployment. Under standard average risk training on mixed ID/OOD distributions, optimization signals from rare OOD events can be overwhelmed by frequent in distribution (ID) patterns, so strong benchmark accuracy may not translate into reliability under high impact shifts. To address this issue, we propose VLBM (Variational Latent Basis Model), a theory guided latent forecasting framework that separates stable dynamics from OOD induced deviations. VLBM learns a shared latent basis that defines a low rank subspace for stable ID dynamics, explicitly decomposes inputs into basis subspace components and orthogonal residual components, and aligns a future aware posterior with a future blind prior so that test time latent inference depends only on historical input. Across 12 benchmark tasks spanning transportation, weather, power systems, and other real world domains, including newly constructed real world OOD traffic datasets, VLBM achieves state of the art OOD robustness and ID accuracy, with average MAE and MSE gains of 15.08\% and 7.74\% over the strongest baseline. On a synthetic simulation dataset, VLBM also consistently achieves the best performance and better tracks OOD pulse recovery. These results support latent structured forecasting as a principled route to robust prediction under mixed ID and OOD conditions. The code is available at https://github.com/leijieruilq/VLBM_OOD_forecast.
