Deep, Skinny Neural Networks are not Universal Approximators.
Jesse JohnsonPublished in: ICLR (Poster) (2019)
Keyphrases
- information systems
- neural network
- multi layer
- pattern recognition
- neural nets
- fault diagnosis
- genetic algorithm
- artificial neural networks
- feed forward
- hopfield networks
- databases
- back propagation
- fuzzy logic
- rough sets
- computational intelligence
- neural network model
- multilayer perceptron
- recurrent neural networks
- expert systems
- computer vision
- fuzzy systems
- training process
- multi layer perceptron
- feedforward neural networks
- competitive learning
- auto associative
- belief nets
- real world