arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2402.13077cs.LGcs.AIcs.NE
收藏

面向科学机器学习的机制神经网络

Mechanistic Neural Networks for Scientific Machine Learning

University of Amsterdam(阿姆斯特丹大学);Institute of Science and Technology(科技研究院)

Adeel Pervez, Francesco Locatello, Efstratios Gavves

AI总结:

本文提出机制神经网络,通过机制模块显式学习控制微分方程表征,并引入NeuRLP松弛线性规划求解器突破传统ODE求解限制,在方程发现和动态系统建模等科学机器学习任务中展现出优于专用最优方法的性能。

摘要:

本文提出了机制神经网络(Mechanistic Neural Networks),这是一种面向科学领域机器学习应用的神经网络设计。它在标准架构中引入了一种新的机制模块(Mechanistic Block),以显式学习作为表征的控制微分方程,从而揭示数据背后的动力学规律,并提升数据建模的可解释性和效率。我们方法的核心是一种新颖的松弛线性规划求解器(NeuRLP),其灵感来源于一种将线性常微分方程求解转化为线性规划求解的技术。该求解器能够与神经网络良好集成,并突破了传统常微分方程求解器的局限,支持可扩展的GPU并行处理。总体而言,机制神经网络展示了其在科学机器学习应用中的广泛适用性,能够熟练处理从方程发现到动态系统建模等各类任务。我们证明了其在分析和解释各种应用中复杂科学数据方面的综合能力,并在与专门的当前最优方法的比较中表现出显著性能。

英文摘要:

This paper presents Mechanistic Neural Networks, a neural network design for machine learning applications in the sciences. It incorporates a new Mechanistic Block in standard architectures to explicitly learn governing differential equations as representations, revealing the underlying dynamics of data and enhancing interpretability and efficiency in data modeling. Central to our approach is a novel Relaxed Linear Programming Solver (NeuRLP) inspired by a technique that reduces solving linear ODEs to solving linear programs. This integrates well with neural networks and surpasses the limitations of traditional ODE solvers enabling scalable GPU parallel processing. Overall, Mechanistic Neural Networks demonstrate their versatility for scientific machine learning applications, adeptly managing tasks from equation discovery to dynamic systems modeling. We prove their comprehensive capabilities in analyzing and interpreting complex scientific data across various applications, showing significant performance against specialized state-of-the-art methods.

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。

↑