Blended Coarse Gradient Descent for Full Quantization of Deep Neural Networks.
Penghang YinShuai ZhangJiancheng LyuStanley J. OsherYingyong QiJack XinPublished in: CoRR (2018)
Keyphrases
- neural network
- cost function
- learning rules
- quantization error
- back propagation
- artificial neural networks
- pattern recognition
- coarse to fine
- genetic algorithm
- multiresolution
- objective function
- loss function
- feed forward
- competitive learning
- network architecture
- fuzzy systems
- blended learning
- online learning
- multilayer perceptron
- training algorithm
- neural network model
- activation function
- radial basis function
- error function
- case study
- deep learning
- conjugate gradient