Learnable Companding Quantization for Accurate Low-bit Neural Networks.
Kohei YamamotoPublished in: CoRR (2021)
Keyphrases
- neural network
- pattern recognition
- highly accurate
- genetic algorithm
- uniform quantization
- multi layer
- fuzzy logic
- artificial neural networks
- high quality
- computationally efficient
- back propagation
- quantization error
- adaptive quantization
- fault diagnosis
- neural network model
- feed forward
- multilayer perceptron
- neural nets
- pac learning
- learning algorithm