On improvements of neural network accuracy with fixed number of active neurons.
Natalia SokolovaDmitry P. NikolaevDmitry PolevoyPublished in: ICMV (2014)
Keyphrases
- fixed number
- neural network
- hidden layer
- high accuracy
- associative memory
- back propagation
- training algorithm
- finite number
- prediction accuracy
- fault diagnosis
- activation function
- neural network model
- artificial neural networks
- computational cost
- neural model
- identical parallel machines
- neural nets
- feedforward neural networks
- error rate
- fuzzy logic
- connection weights
- neuron model
- spiking neurons
- fuzzy neural network
- network architecture
- learning rate
- recurrent neural networks
- feed forward
- radial basis function
- lower bound
- support vector
- genetic algorithm