Gradient-enhanced physics-informed neural networks based on transfer learning for inverse problems of the variable coefficient differential equations.
Shuning LinYong ChenPublished in: CoRR (2023)
Keyphrases
- differential equations
- transfer learning
- inverse problems
- neural network
- partial differential equations
- image reconstruction
- global optimization
- convex optimization
- machine learning
- reinforcement learning
- dynamical systems
- pattern recognition
- optimization methods
- labeled data
- image denoising
- active learning
- optimization problems
- semi supervised learning
- cross domain
- back propagation
- image processing
- artificial neural networks
- collaborative filtering
- early vision
- text classification
- image enhancement
- learning algorithm
- structure learning
- transfer knowledge
- level set
- pairwise
- multiscale
- computer vision
- similarity measure
- smoothness constraint
- machine learning algorithms
- object recognition
- super resolution