Neural network identifiability for a family of sigmoidal nonlinearities.
Verner VlacicHelmut BölcskeiPublished in: CoRR (2019)
Keyphrases
- neural network
- feed forward neural networks
- activation function
- artificial neural networks
- feedforward neural networks
- back propagation
- pattern recognition
- special case
- neural nets
- network architecture
- neural network is trained
- hybrid intelligent
- neural network model
- feed forward
- hidden layer
- image reconstruction from projections
- nonlinear functions
- training algorithm
- multi layer
- multilayer perceptron
- basis functions
- self organizing maps
- artificial intelligence
- recurrent neural networks
- fuzzy neural network
- fuzzy logic
- causal discovery
- training set
- data sets