Homotopy-based training of NeuralODEs for accurate dynamics discovery.
Joon-Hyuk KoHankyul KohNojun ParkWonho JhePublished in: CoRR (2022)
Keyphrases
- recurrent networks
- high accuracy
- dynamic model
- high quality
- training set
- training phase
- training process
- training samples
- computationally efficient
- data mining
- scientific discovery
- high precision
- supervised learning
- artificial intelligence
- real world
- data sets
- text classification
- artificial neural networks
- data structure
- dynamical systems
- neural network
- highly accurate
- database systems
- highly nonlinear
- databases