Arbitrary-Depth Universal Approximation Theorems for Operator Neural Networks.
Annan YuChloé BecqueyDiana HalikiasMatthew Esmaili MalloryAlex TownsendPublished in: CoRR (2021)
Keyphrases
- neural network
- artificial neural networks
- back propagation
- pattern recognition
- multi valued
- recurrent neural networks
- upper approximation
- continuous functions
- competitive learning
- approximation algorithms
- neural network model
- depth map
- three dimensional
- fuzzy systems
- network architecture
- rule extraction
- relative error
- approximation error
- genetic algorithm
- approximation methods
- data sets
- error bounds
- sufficient conditions
- np hard
- approximation schemes
- laplacian operator