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 2308.06338cs.LGcs.CCcs.NAmath.APmath.NA
收藏

深度算子网络的大小下界

Size Lowerbounds for Deep Operator Networks

The University of Manchester(曼彻斯特大学);Robert Bosch GmbH(罗伯特·博世有限公司);Jadavpur University(贾达普大学)

Anirbit Mukherjee, Amartya Roy

AI总结:

本文首次为深度算子网络建立了数据依赖型规模下界,证明在 n 个数据点上获得低训练误差需要公共输出维度按 Ω(√[4]{n}) 增长,并通过求解对流-扩散-反应 PDE 的实验表明训练数据规模可能需随该维度至少二次方增长。

摘要:

深度算子网络(Deep Operator Networks)是一种日益流行的范式,用于求解无穷维空间中的回归问题,从而一次性求解偏微分方程族。在本工作中,我们旨在建立首个数据依赖型的下界,用以刻画深度算子网络(DeepONets)在噪声数据上降低经验误差所需的最小规模。特别地,我们证明:若要在 n 个数据点上获得较低的训练误差,分支网络(branch net)与主干网络(trunk net)的公共输出维度必须按 Ω(√[4]{n}) 的规模增长。这一结果启发了我们使用深度算子网络求解对流-扩散-反应偏微分方程的实验。在实验中,我们展示了这样一种可能性:在固定模型规模下,若要利用公共输出维度的增加来获得训练误差的单调下降,训练数据规模可能至少需要随该维度呈二次方增长。

英文摘要:

Deep Operator Networks are an increasingly popular paradigm for solving regression in infinite dimensions and hence solve families of PDEs in one shot. In this work, we aim to establish a first-of-its-kind data-dependent lowerbound on the size of DeepONets required for them to be able to reduce empirical error on noisy data. In particular, we show that for low training errors to be obtained on $n$ data points it is necessary that the common output dimension of the branch and the trunk net be scaling as $Ω\left ( \sqrt[\leftroot{-1}\uproot{-1}4]{n} \right )$. This inspires our experiments with DeepONets solving the advection-diffusion-reaction PDE, where we demonstrate the possibility that at a fixed model size, to leverage increase in this common output dimension and get monotonic lowering of training error, the size of the training data might necessarily need to scale at least quadratically with it.

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。

↑