On sensitivity of neural adaptive filters with respect to the slope parameter of a neuron activation function.
Warren SherlikerIgor R. KrcmarMilorad M. BozicDanilo P. MandicPublished in: ICASSP (2002)
Keyphrases
- activation function
- network architecture
- neural network
- single neuron
- artificial neural networks
- feed forward
- feed forward neural networks
- back propagation
- hidden layer
- neural nets
- learning rate
- chaotic neural network
- network size
- basis functions
- sigmoid function
- neural model
- multilayer perceptron
- training phase
- learning rules
- recurrent neural networks
- neural network model
- hidden nodes
- radial basis function
- genetic algorithm
- prediction model
- feedforward neural networks
- extreme learning machine
- spike trains
- fuzzy logic