On the Importance of Consistency in Training Deep Neural Networks.
Chengxi YeYezhou YangCornelia FermüllerYiannis AloimonosPublished in: CoRR (2017)
Keyphrases
- neural network
- training process
- training algorithm
- feedforward neural networks
- feed forward neural networks
- multi layer perceptron
- pattern recognition
- backpropagation algorithm
- back propagation
- relative importance
- neural network training
- training set
- deep architectures
- supervised learning
- training patterns
- feature space
- fuzzy logic
- recurrent networks
- consistency checking
- recurrent neural networks
- neural network structure
- multilayer perceptron
- error back propagation
- database
- fault diagnosis
- constraint satisfaction problems
- online learning
- artificial neural networks
- training data
- information systems
- real time