Numerical solution for high-dimensional partial differential equations based on deep learning with residual learning and data-driven learning.
Zheng WangFutian WengJialin LiuKai CaoMuzhou HouJuan WangPublished in: Int. J. Mach. Learn. Cybern. (2021)
Keyphrases
- numerical solution
- partial differential equations
- deep learning
- driven learning
- prior knowledge
- active learning
- high dimensional
- data points
- learning algorithm
- image processing
- image denoising
- level set
- training data
- finite element
- numerical methods
- supervised learning
- model selection
- unsupervised learning
- reinforcement learning
- multiscale
- bayesian networks
- image segmentation