Automatic Differentiation is Essential in Training Neural Networks for Solving Differential Equations.
Chuqi ChenYahong YangYang XiangWenrui HaoPublished in: CoRR (2024)
Keyphrases
- differential equations
- neural network
- training process
- feed forward artificial neural networks
- runge kutta
- dynamical systems
- feedforward neural networks
- feed forward neural networks
- training algorithm
- ordinary differential equations
- back propagation
- artificial neural networks
- boundary value problem
- numerical solution
- brownian motion
- fuzzy logic
- training set
- pattern recognition
- neural network model
- numerical methods
- multi layer perceptron
- difference equations
- recurrent neural networks
- natural images