arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2502.02472stat.MLcs.LG
收藏

SDE Matching:潜在随机微分方程的可扩展且免仿真训练

SDE Matching: Scalable and Simulation-Free Training of Latent Stochastic Differential Equations

University of Amsterdam(阿姆斯特丹大学);Constructor University, Bremen(布雷梅构造大学)

Grigory Bartosh, Dmitry Vetrov, Christian A. Naesseth

AI总结:

本文提出SDE Matching,一种免仿真的Latent SDE训练新方法,通过扩展Score-和Flow Matching算法至随机动力学领域,消除了昂贵的数值仿真需求,在保持与伴随灵敏度方法相当性能的同时大幅降低了计算复杂度。

摘要:

潜在随机微分方程(SDE)是时间序列和序列建模的强大工具。然而,训练 Latent SDE 通常依赖于伴随灵敏度方法,该方法依赖于仿真和通过近似 SDE 解的反向传播,这限制了可扩展性。在这项工作中,我们提出了 SDE Matching,一种用于训练 Latent SDE 的新的免仿真方法。受用于学习生成动力学的现代 Score- 和 Flow Matching 算法的启发,我们将这些思想扩展到用于时间序列和序列建模的随机动力学领域,消除了对昂贵的数值仿真的需求。我们的结果表明,SDE Matching 实现了与伴随灵敏度方法相当的性能,同时大幅降低了计算复杂度。

英文摘要:

The Latent Stochastic Differential Equation (SDE) is a powerful tool for time series and sequence modeling. However, training Latent SDEs typically relies on adjoint sensitivity methods, which depend on simulation and backpropagation through approximate SDE solutions, which limit scalability. In this work, we propose SDE Matching, a new simulation-free method for training Latent SDEs. Inspired by modern Score- and Flow Matching algorithms for learning generative dynamics, we extend these ideas to the domain of stochastic dynamics for time series and sequence modeling, eliminating the need for costly numerical simulations. Our results demonstrate that SDE Matching achieves performance comparable to adjoint sensitivity methods while drastically reducing computational complexity.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑