Some theorems concerning the free energy of (un)constrained stochastic hopfield neural networks.
Jan van den BergJan C. BiochPublished in: EuroCOLT (1995)
Keyphrases
- hopfield neural network
- free energy
- energy function
- belief propagation
- cellular neural networks
- fixed point
- upper bound
- energy minimization
- competitive learning
- posterior distribution
- exponential stability
- genetic algorithm
- approximate inference
- learning rules
- graphical models
- sufficient conditions
- interior point methods
- graph cuts
- markov random field
- higher order
- learning algorithm