arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2306.07526cs.LGcs.AI
收藏

面向物理动力系统的扩散模型中用户自定义事件采样与不确定性量化

User-defined Event Sampling and Uncertainty Quantification in Diffusion Models for Physical Dynamical Systems

New York University(纽约大学);Google Research(谷歌研究院)

Marc Finzi, Anudhyan Boral, Andrew Gordon Wilson, Fei Sha, Leonardo Zepeda-Núñez

AI总结:

本研究针对扩散模型在混沌动力系统应用中难以查询隐式极端事件知识的问题,提出条件得分函数的概率近似方案,可实现推理阶段非线性用户自定义事件的条件采样并匹配数据统计特性。

摘要:

扩散模型是一类概率生成模型,已被广泛用作图像处理任务(如文本条件生成和图像修复)的先验。我们证明,这些模型可经调整用于对混沌动力系统进行预测并提供不确定性量化。在这类应用中,扩散模型能够隐式表征关于异常值和极端事件的知识;然而,通过条件采样或概率测量来查询这类知识却异常困难。现有的推理阶段条件采样方法主要致力于施加约束,这不足以匹配分布的统计特性,也无法计算所选事件的概率。要实现这些目标,最优方式是使用条件得分函数,但其计算通常是难以处理的。在本研究中,我们提出了一种条件得分函数的概率近似方案,该方案可证明随着噪声水平降低会收敛到真实分布。借助该方案,我们能够在推理阶段基于非线性用户自定义事件进行条件采样,且即使从分布的尾部采样,也能匹配数据统计特性。

英文摘要:

Diffusion models are a class of probabilistic generative models that have been widely used as a prior for image processing tasks like text conditional generation and inpainting. We demonstrate that these models can be adapted to make predictions and provide uncertainty quantification for chaotic dynamical systems. In these applications, diffusion models can implicitly represent knowledge about outliers and extreme events; however, querying that knowledge through conditional sampling or measuring probabilities is surprisingly difficult. Existing methods for conditional sampling at inference time seek mainly to enforce the constraints, which is insufficient to match the statistics of the distribution or compute the probability of the chosen events. To achieve these ends, optimally one would use the conditional score function, but its computation is typically intractable. In this work, we develop a probabilistic approximation scheme for the conditional score function which provably converges to the true distribution as the noise level decreases. With this scheme we are able to sample conditionally on nonlinear userdefined events at inference time, and matches data statistics even when sampling from the tails of the distribution.

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。

↑