arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2202.08494cs.LG
收藏

为连续物理学习连续模型

Learning continuous models for continuous physics

International Computer Science Institute(国际计算机科学研究所);Lawrence Berkeley National Laboratory(劳伦斯伯克利国家实验室);University of California Berkeley(加州大学伯克利分校);Google Research(谷歌研究院)

Aditi S. Krishnapriyan, Alejandro F. Queiruga, N. Benjamin Erichson, Michael W. Mahoney

AI总结:

针对机器学习模型在离散数据上训练导致无法捕捉底层连续动力学的问题,基于数值分析理论开发了一种收敛性测试,以验证模型对连续动力学的近似准确性,从而提升科学预测任务中的插值与外推能力。

摘要:

随时间连续演化的动力系统在科学和工程中无处不在。机器学习(ML)提供了数据驱动的方法来建模和预测此类系统的动力学。该方法的一个核心问题是,ML模型通常在离散数据上训练,使用的ML方法并未感知底层的连续性性质。这导致模型往往无法捕捉任何底层连续动力学——无论是目标系统的,还是任何相关系统的。为应对这一挑战,我们基于数值分析理论开发了一种收敛性测试。我们的测试验证模型是否学习到了一个能准确近似底层连续动力学的函数。未通过此测试的模型无法捕捉相关动力学,使其在许多科学预测任务中效用有限;而通过此测试的模型则能在多种方面实现更好的插值和更好的外推。我们的结果说明了如何将有原则的数值分析方法与现有的ML训练/测试方法相结合,以验证用于科学和工程应用的模型。

英文摘要:

Dynamical systems that evolve continuously over time are ubiquitous throughout science and engineering. Machine learning (ML) provides data-driven approaches to model and predict the dynamics of such systems. A core issue with this approach is that ML models are typically trained on discrete data, using ML methodologies that are not aware of underlying continuity properties. This results in models that often do not capture any underlying continuous dynamics -- either of the system of interest, or indeed of any related system. To address this challenge, we develop a convergence test based on numerical analysis theory. Our test verifies whether a model has learned a function that accurately approximates an underlying continuous dynamics. Models that fail this test fail to capture relevant dynamics, rendering them of limited utility for many scientific prediction tasks; while models that pass this test enable both better interpolation and better extrapolation in multiple ways. Our results illustrate how principled numerical analysis methods can be coupled with existing ML training/testing methodologies to validate models for science and engineering applications.

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。

↑