BinaryConnect: Training Deep Neural Networks with binary weights during propagations.
Matthieu CourbariauxYoshua BengioJean-Pierre DavidPublished in: NIPS (2015)
Keyphrases
- neural network
- training process
- training algorithm
- feedforward neural networks
- neural network training
- multi layer perceptron
- artificial neural networks
- back propagation
- training patterns
- feed forward neural networks
- feedforward artificial neural networks
- pattern recognition
- training examples
- neural network structure
- deep architectures
- hidden neurons
- training set
- backpropagation algorithm
- connection weights
- self organizing maps
- training samples
- linear combination
- fuzzy logic
- error back propagation
- activation function
- weighted sum
- linearly combined
- weight update
- neural network model
- output layer
- online training
- supervised learning
- number of hidden units
- hidden units
- deep learning
- non binary
- rbf network
- recurrent neural networks
- fault diagnosis
- decision trees