Expectation Backpropagation: Parameter-Free Training of Multilayer Neural Networks with Continuous or Discrete Weights.
Daniel SoudryItay HubaraRon MeirPublished in: NIPS (2014)
Keyphrases
- parameter free
- multilayer neural networks
- back propagation
- learning algorithm
- learning rate
- training algorithm
- bp algorithm
- hidden layer
- backpropagation algorithm
- feedforward neural networks
- feed forward neural networks
- connection weights
- neural network
- single hidden layer
- feed forward
- weight update
- artificial neural networks
- multilayer perceptron neural network
- activation function
- outlier detection
- categorical data
- error back propagation
- multilayer neural network
- multilayer perceptron
- neural nets
- extreme learning machine
- fuzzy logic
- backpropagation neural networks
- bp neural network
- backpropagation neural network
- convergence rate
- fully automatic
- neural network ensemble
- linear combination
- training set
- supervised learning
- artificial intelligence
- hidden nodes
- data mining techniques
- recurrent neural networks
- detection algorithm
- machine learning