Study of Asymptotic Behavior of a Broad Class of Criteria for Data-Driven Best Model Selection.
Volodymyr StepashkoPublished in: CSIT (2020)
Keyphrases
- model selection
- data driven
- cross validation
- model selection criteria
- sample size
- marginal likelihood
- parameter estimation
- hyperparameters
- machine learning
- statistical learning
- selection criterion
- bayesian learning
- motion segmentation
- feature selection
- selection criteria
- learning machines
- gaussian process
- variable selection
- regression model
- meta learning
- generalization bounds
- error estimation
- bayesian information criterion
- mixture model
- bayesian networks
- parameter determination
- bayesian model selection
- automatic model selection
- multivariate regression