Improving Binary Neural Networks through Fully Utilizing Latent Weights.
Weixiang XuQiang ChenXiangyu HePeisong WangJian ChengPublished in: CoRR (2021)
Keyphrases
- neural network
- pattern recognition
- latent variables
- linear combination
- genetic algorithm
- artificial neural networks
- feed forward
- weight update
- neural network model
- back propagation
- multilayer perceptron
- weighted sum
- hidden neurons
- multi layer
- hamming distance
- weighting scheme
- non binary
- machine learning
- fault diagnosis
- constraint satisfaction problems
- activation function
- weighted average
- optimal weights