Deep-learning PDEs with unlabeled data and hardwiring physics laws
This provides a fast and accessible solution for scientific disciplines dealing with PDEs, though it builds incrementally on earlier works for linear and homogeneous cases.
The paper tackles solving nonlinear and non-homogeneous partial differential equations using convolutional deep neural networks without labeled data, achieving accurate results as demonstrated in their experiments.
Providing fast and accurate solutions to partial differential equations is a problem of continuous interest to the fields of applied mathematics and physics. With the recent advances in machine learning, the adoption learning techniques in this domain is being eagerly pursued. We build upon earlier works on linear and homogeneous PDEs, and develop convolutional deep neural networks that can accurately solve nonlinear and non-homogeneous equations without the need for labeled data. The architecture of these networks is readily accessible for scientific disciplines who deal with PDEs and know the basics of deep learning.