Mean-field theory of two-layers neural networks: dimension-free bounds and kernel limit.
Song MeiTheodor MisiakiewiczAndrea MontanariPublished in: CoRR (2019)
Keyphrases
- neural network
- multi layer
- lower bound
- pattern recognition
- artificial neural networks
- upper bound
- single layer
- kernel function
- genetic algorithm
- feature space
- back propagation
- lower and upper bounds
- neural nets
- fuzzy systems
- support vector
- multiple kernel learning
- upper and lower bounds
- kernel machines
- error bounds
- multilayer perceptron
- neural network model
- kernel methods
- worst case
- belief networks
- gaussian processes
- kernel matrix
- self organizing maps
- markov random field
- feature selection
- multiple layers