Academy of Mathematics and Systems Science, CAS
Colloquia & Seminars

Speaker:	周涛研究员，计算与科学工程计算数学研究所
Inviter:
Title:	Deep adaptive sampling for numerical PDEs
Time & Venue:	2022.11.11 11:20-12:00 南楼204 腾讯会议:553-730-880
Abstract:	Adaptive computation is of great importance in numerical simulations. The ideas for adaptive computations can be dated back to adaptive finite element methods in 1970s. In this talk, we shall first review some recent development for adaptive method with applications. Then, we shall propose a deep adaptive sampling method for solving PDEs where deep neural networks are utilized to approximate the solutions. In particular, we propose the failure informed PINNs (FI-PINNs), which can adaptively refine the training set with the goal of reducing the failure probability. Compared to the neural network approximation obtained with uniformly distributed collocation points, the developed algorithms can significantly improve the accuracy, especially for low regularity and high-dimensional problems.