Sharp asymptotics on the compression of two-layer neural networks.
Mohammad Hossein AmaniSimone BombariMarco MondelliRattana PukdeeStefano RiniPublished in: ITW (2022)
Keyphrases
- neural network
- multi layer
- single layer
- pattern recognition
- artificial neural networks
- image compression
- sufficient conditions
- compression ratio
- compression scheme
- data compression
- compression algorithm
- back propagation
- neural nets
- neural network model
- multilayer perceptron
- competitive learning
- multiple layers
- fuzzy systems
- feed forward
- application layer
- recurrent neural networks
- large deviations
- image processing
- middle layer
- heavy traffic
- lossy compression
- activation function
- fault diagnosis
- markov chain
- fuzzy logic
- machine learning