Pre-training of Recurrent Neural Networks via Linear Autoencoders.
Luca PasaAlessandro SperdutiPublished in: NIPS (2014)
Keyphrases
- recurrent neural networks
- feedforward neural networks
- recurrent networks
- echo state networks
- hidden layer
- feed forward
- neural network
- artificial neural networks
- complex valued
- feed forward neural networks
- reservoir computing
- denoising
- training algorithm
- output layer
- neural model
- long short term memory
- training process
- closed form
- training set
- nonlinear dynamic systems
- cascade correlation
- real time
- linear svm
- training samples
- restricted boltzmann machine
- long term
- hebbian learning
- artificial intelligence
- machine learning