Network Model Selection Using Task-Focused Minimum Description Length.
Ivan BrugereTanya Y. Berger-WolfPublished in: WWW (Companion Volume) (2018)
Keyphrases
- model selection
- minimum description length
- cross validation
- parameter estimation
- hyperparameters
- mixture model
- mdl principle
- sample size
- error estimation
- bayesian learning
- machine learning
- meta learning
- feature selection
- selection criterion
- statistical learning
- motion segmentation
- information theoretic
- model selection criteria
- generalization error
- marginal likelihood
- statistical inference
- regression model
- parameter determination
- image processing
- network structure
- gaussian process
- bayesian methods
- image segmentation
- generalization bounds
- bayesian information criterion
- decision trees
- information theory
- variable selection