BinaryNet: Training Deep Neural Networks with Weights and Activations Constrained to +1 or -1.
Matthieu CourbariauxYoshua BengioPublished in: CoRR (2016)
Keyphrases
- neural network
- training process
- training algorithm
- feedforward neural networks
- hidden neurons
- feed forward neural networks
- artificial neural networks
- multi layer perceptron
- back propagation
- training speed
- feedforward artificial neural networks
- linear combination
- test set
- neural network training
- recurrent neural networks
- training patterns
- training data
- multi layer
- error back propagation
- connection weights
- weight update
- genetic algorithm
- neural network structure
- training phase
- relative importance
- weighted sum
- multilayer perceptron
- training set
- pattern recognition
- fuzzy neural network
- network architecture
- neural nets
- feed forward
- supervised learning
- deep architectures
- weighted sums
- support vector
- number of hidden units