Local conditional and marginal approach to parameter estimation in discrete graphical models.
Hélène MassamNanwei WangPublished in: J. Multivar. Anal. (2018)
Keyphrases
- parameter estimation
- graphical models
- approximate inference
- chain graphs
- structure learning
- belief propagation
- maximum likelihood
- probabilistic model
- least squares
- random variables
- em algorithm
- markov random field
- probabilistic graphical models
- graphical structure
- statistical models
- bayesian networks
- model selection
- expectation maximization
- probabilistic inference
- conditional random fields
- markov networks
- exact inference
- factor graphs
- parameter estimation algorithm
- belief networks
- probability distribution
- random fields
- conditional probabilities
- conditional independence
- maximum likelihood estimation
- markov chain monte carlo
- variational methods
- posterior distribution
- marginal distributions
- statistical relational learning
- partition function
- high dimensional