Selection of classifiers based on the MDL principle using the VC dimension.
Mineichi KudoMasaru ShimboPublished in: ICPR (1996)
Keyphrases
- vc dimension
- mdl principle
- generalization bounds
- minimum description length
- learning machines
- upper bound
- sample size
- lower bound
- risk bounds
- concept classes
- sample complexity
- information theory
- inductive inference
- statistical learning theory
- learning algorithm
- worst case
- empirical risk minimization
- decision trees
- concept class
- pac learning
- support vector
- feature selection
- training examples
- training data
- supervised classification
- linear classifiers
- information theoretic
- generalization ability
- semi supervised learning
- semi supervised
- feature extraction