Physics-Informed Neural Networks (PINNs) for Parameterized PDEs: A Metalearning Approach.
Michael PenwardenShandian ZheAkil NarayanRobert M. KirbyPublished in: CoRR (2021)
Keyphrases
- neural network
- pattern recognition
- partial differential equations
- artificial neural networks
- meta learning
- genetic algorithm
- image processing
- level set
- back propagation
- feed forward
- anisotropic diffusion
- data mining problems
- computer science
- fuzzy systems
- fuzzy logic
- artificial intelligence
- neural nets
- rule extraction
- training process
- network architecture
- image denoising
- diffusion equation
- fourth order
- multi layer
- recurrent neural networks
- high order
- multistage
- sufficient conditions
- edge detection
- expert systems
- computer vision
- data mining