ATA: Attentional Non-Linear Activation Function Approximation for VLSI-Based Neural Networks.
Linyu WeiJueping CaiWuzhuang WangPublished in: IEEE Signal Process. Lett. (2021)
Keyphrases
- activation function
- neural network
- continuous functions
- feed forward
- feed forward neural networks
- hidden layer
- artificial neural networks
- back propagation
- feedforward neural networks
- hidden neurons
- single hidden layer
- backpropagation algorithm
- network architecture
- multilayer perceptron
- hidden nodes
- learning rate
- neural nets
- pattern recognition
- chaotic neural network
- neural network model
- radial basis function
- fuzzy neural network
- training phase
- extreme learning machine
- basis functions
- single layer
- fault diagnosis
- signal processing
- hidden units
- fuzzy logic
- recurrent neural networks
- autoregressive
- rbf neural network
- knn