Interweaving Marginals into Multivariate Sample Paths: Training-Free Dependence Construction for Probabilistic Time Series Foundation Models

2026-09-22

AI Summary

研究通过无训练方法将冻结的时间序列基础模型边际结合成多变量预测样本路径，以改善依赖关系诊断。

Abstract

Probabilistic time series foundation models (TSFMs) provide coordinate-wise predictive distributions, but these marginals do not determine a joint distribution over multivariate future trajectories. We study training-free coupling of frozen TSFM marginals into multivariate forecast sample paths. Our primary evaluation fixes the empirical marginal sample multiset at every channel--horizon coordinate across methods, isolating the effect of coupling alone. Historical temporal and channel relations substantially improve their corresponding dependence diagnostics. The same pattern persists when the fixed-marginal constraint is removed and paths are sampled directly, and remains present under native multivariate backbone inference. These results support treating dependence reconstruction as a distinct post-processing problem for probabilistic TSFMs.

Problem

• Probabilistic time series foundation models
• Multivariate forecast sample paths
• Training-free coupling
• Marginal distributions
• Joint distribution

Innovation

• training-free coupling
• frozen TSFM marginals
• multivariate forecast sample paths
• dependence reconstruction