Range-Lookup Approximate Computing Acceleration for Any Activation Functions in Low-Power Neural Network.
Wen-Chang YangShu-Yun LinTsung-Chu HuangPublished in: ICCE-TW (2020)
Keyphrases
- low power
- activation function
- neural network
- high speed
- power consumption
- low cost
- feed forward neural networks
- back propagation
- hidden layer
- feed forward
- artificial neural networks
- neural architecture
- hidden neurons
- multilayer perceptron
- network architecture
- neural nets
- connection weights
- low power consumption
- feedforward neural networks
- vlsi circuits
- fuzzy neural network
- learning rate
- hidden nodes
- neural network model
- radial basis function
- sigmoid function
- recurrent neural networks
- basis functions
- multi layer perceptron
- logic circuits
- expert systems
- gate array
- rbf neural network
- cmos technology
- autoregressive
- multi layer
- bp neural network
- parallel processing
- fault diagnosis
- fuzzy logic
- pattern recognition
- feature selection