A Penalization Criterion Based on Noise Behaviour for Model Selection.
Joaquín Pizarro JunqueraPedro Galindo RiañoElisa Guerrero VázquezAndrés Yáñez EscolanoPublished in: IWANN (2) (2001)
Keyphrases
- model selection
- selection criterion
- model selection criteria
- generalization error
- cross validation
- marginal likelihood
- variable selection
- information criterion
- feature selection
- hyperparameters
- regression model
- bayesian learning
- statistical learning
- bayesian information criterion
- sample size
- bayesian approaches
- parameter estimation
- error estimation
- motion segmentation
- machine learning
- noise level
- automatic model selection
- missing data
- meta learning
- mixture model
- gaussian process
- statistical inference
- bayesian methods
- leave one out cross validation
- active learning
- image segmentation