arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1906.02160cs.LGcs.ROphysics.comp-phstat.ML
收藏

变分高斯过程的物理增强数据驱动模型

Physics Enhanced Data-Driven Models with Variational Gaussian Processes

Virginia Commonwealth University(弗吉尼亚联邦大学)

Daniel L. Marino, Milos Manic

摘要:

自然科学和数学建模几个世纪的发展提供了宝贵的领域专家知识,这些知识尚未被探索用于机器学习模型的开发。在对复杂的物理系统建模时,领域知识和数据都提供了关于系统的必要信息。在本文中,我们提出了一个数据驱动的模型,利用部分领域知识,以提高泛化和可解释性。我们称之为EVGP(Explicit Variational GaussianProcess)的方法具有以下优点:1)使用可用的领域知识来改进模型的假设(归纳偏差),2)可扩展到大数据集,3)提高可解释性。我们展示了如何EVGP模型可以用来学习系统动力学使用基本的牛顿力学作为先验知识。我们演示了如何将先验领域知识添加到数据驱动模型中,从而优于纯数据驱动模型。

英文摘要:

Centuries of development in natural sciences and mathematical modeling provide valuable domain expert knowledge that has yet to be explored for the development of machine learning models. When modeling complex physical systems, both domain knowledge and data provide necessary information about the system. In this paper, we present a data-driven model that takes advantage of partial domain knowledge in order to improve generalization and interpretability. The presented approach, which we call EVGP (Explicit Variational GaussianProcess), has the following advantages: 1) using available domain knowledge to improve the assumptions(inductive bias) of the model, 2) scalability to large datasets, 3) improved interpretability. We show how the EVGP model can be used to learn system dynamics using basic Newtonian mechanics as prior knowledge. We demonstrate how the addition of prior domain-knowledge to data-driven models outperforms purely data-driven models.

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。

↑