Maximum Likelihood Model Selection for 1-Norm Soft Margin SVMs with Multiple Parameters.
Tobias GlasmachersChristian IgelPublished in: IEEE Trans. Pattern Anal. Mach. Intell. (2010)
Keyphrases
- model selection
- soft margin
- parameter selection
- hyperparameters
- maximum likelihood
- generalization error
- likelihood function
- parameter estimation
- support vector
- cross validation
- gaussian process
- mixture model
- sample size
- machine learning
- em algorithm
- feature selection
- support vector machine
- bayesian methods
- regression model
- variable selection
- expectation maximization
- binary classification
- maximum a posteriori
- linear classifiers
- gaussian mixture model
- multi class
- ls svm
- svm classifier
- active learning
- boosting algorithms