Incorporating Expressive Graphical Models in Variational Approximations: Chain-Graphs and Hidden Variables
Tal El-HayNir FriedmanPublished in: CoRR (2013)
Keyphrases
- chain graphs
- hidden variables
- graphical models
- belief propagation
- free energy
- models with hidden variables
- variational methods
- approximate inference
- probabilistic model
- bayesian networks
- probabilistic graphical models
- conditional independence
- generative model
- em algorithm
- random variables
- markov networks
- bayesian inference
- latent variables
- probabilistic inference
- structure learning
- image segmentation
- exponential family
- posterior distribution
- belief networks
- message passing
- exact inference
- conditional random fields
- factor graphs
- semi supervised
- structured prediction
- upper bound
- partition function
- undirected graphical models
- reinforcement learning
- graph structure