Eigenvalue Decay Implies Polynomial-Time Learnability for Neural Networks.
Surbhi GoelAdam R. KlivansPublished in: NIPS (2017)
Keyphrases
- neural network
- covariance matrix
- fuzzy logic
- least squares
- pattern recognition
- artificial neural networks
- multilayer perceptron
- feed forward
- multi layer
- self organizing maps
- neural network model
- auto associative
- neuro fuzzy systems
- rule extraction
- fuzzy neural network
- genetic algorithm
- back propagation
- multi agent
- perturbation theory
- machine learning
- competitive learning
- feedforward neural networks
- image processing
- power law
- activation function
- training process
- fuzzy systems
- expert systems
- neural nets
- associative memory
- control system
- computational intelligence
- databases