TL;DR本文从神经切向核角度研究了具有物理约束的神经网络的训练以及其训练过程中收敛率不同的 loss 组件,提出了一种利用 NTK 的特征值来自适应地校准误差收敛率的优化算法。
Abstract
physics-informed neural networks (PINNs) have lately received great attention
thanks to their flexibility in tackling a wide range of forward and inverse
problems involving partial differential equations. However