Post-training Piecewise Linear Quantization for Deep Neural Networks.
Jun FangAli ShafieeHamzah Abdel-AzizDavid ThorsleyGeorgios GeorgiadisJoseph HassounPublished in: ECCV (2) (2020)
Keyphrases
- piecewise linear
- neural network
- training process
- training algorithm
- dynamic programming
- finite sets
- chaotic map
- regression algorithm
- multi layer perceptron
- back propagation
- pattern recognition
- feedforward neural networks
- artificial neural networks
- training examples
- hyperplane
- training set
- deep architectures
- recurrent neural networks
- principal curves
- training samples
- knn
- high resolution
- decision trees
- learning algorithm