Model Selection Method for AdaBoost Using Formal Information Criteria.
Daisuke KajiSumio WatanabePublished in: ICONIP (2) (2008)
Keyphrases
- model selection
- cross validation
- parameter estimation
- hyperparameters
- pairwise
- learning algorithm
- prior knowledge
- information criterion
- error estimation
- sample size
- model selection criteria
- selection criterion
- bayesian learning
- machine learning
- feature selection
- missing data
- similarity measure
- objective function
- closed form
- prior information
- markov random field
- mutual information
- feature set
- multi class
- classification accuracy
- support vector