arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2407.09032math.NAcs.LGcs.NA
收藏

DRM 再探讨:一项完整的误差分析

DRM Revisited: A Complete Error Analysis

School of Mathematics and Statistics, Wuhan University(武汉大学数学与统计学院);Hubei Key Laboratory of Computational Science, Wuhan University(武汉大学湖北计算科学重点实验室);School of Mathematical Sciences, Peking University(北京大学数学科学学院)

Yuling Jiao, Ruoxuan Li, Peiying Wu, Jerry Zhijian Yang, Pingwen Zhang

AI总结:

该研究聚焦过参数化场景下Deep Ritz Method(DRM)的基础理论问题,旨在明确达到目标精度所需的训练样本量、网络架构参数、投影梯度下降步长及迭代次数,以实现对偏微分方程真解的高精度近似,完善DRM的误差分析体系。

摘要:

在本研究中,我们针对过参数化场景下深度里茨法(Deep Ritz Method,DRM)理论分析中的一个基础问题展开研究:给定目标精度水平,如何确定合适的训练样本数量、神经网络的关键架构参数、投影梯度下降优化过程的步长,以及所需的迭代次数,使得梯度下降过程的输出能够以指定精度紧密近似底层偏微分方程的真实解?

英文摘要:

In this work, we address a foundational question in the theoretical analysis of the Deep Ritz Method (DRM) under the over-parameteriztion regime: Given a target precision level, how can one determine the appropriate number of training samples, the key architectural parameters of the neural networks, the step size for the projected gradient descent optimization procedure, and the requisite number of iterations, such that the output of the gradient descent process closely approximates the true solution of the underlying partial differential equation to the specified precision?

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。

↑