Restricted Boltzmann machines for pre-training deep Gaussian networks.
Mark EastwoodChrisina JaynePublished in: IJCNN (2013)
Keyphrases
- restricted boltzmann machine
- deep learning
- deep belief networks
- contrastive divergence
- conditional random fields
- probabilistic graphical models
- learning algorithm
- unsupervised learning
- semi supervised
- machine learning
- hidden layer
- loopy belief propagation
- markov random field
- graph cuts
- sparse coding
- artificial neural networks
- pairwise
- training set