A two phase approach to Bayesian network model selection and comparison between the MDL and DGM scoring heuristics.
Michael J. KaneFerat SahinAndreas E. SavakisPublished in: SMC (2003)
Keyphrases
- model selection
- bayesian networks
- cross validation
- hyperparameters
- parameter estimation
- sample size
- statistical learning
- bayesian learning
- error estimation
- regression model
- generalization error
- parameter determination
- selection criterion
- machine learning
- mixture model
- gaussian process
- statistical inference
- graphical models
- model selection criteria
- feature selection
- variable selection
- posterior distribution
- information criterion
- marginal likelihood
- causal relationships
- probabilistic model
- leave one out cross validation
- learning bayesian networks
- bayesian information criterion