Maximum margin structure learning of Bayesian network classifiers.
Franz PernkopfMichael WohlmayrManfred MückePublished in: ICASSP (2011)
Keyphrases
- bayesian network classifiers
- structure learning
- maximum margin
- markov networks
- graphical models
- bayesian networks
- belief propagation
- structured prediction
- maximum likelihood
- discriminative learning
- first order logic
- parameter learning
- support vector
- bayesian inference
- probabilistic model
- hyperplane
- transfer learning
- parameter estimation
- conditional random fields
- hidden variables
- document classification
- learning algorithm
- approximate inference
- markov random field
- conditional independence
- multi task
- sample size
- generalization error
- pattern classification
- generative model
- posterior probability
- higher order
- cross validation
- stereo matching
- network structure