Selection of Classifiers Using Information-Theoretic Criteria.
Hee-Joong KangPublished in: ICAPR (1) (2005)
Keyphrases
- information theoretic
- information theory
- mutual information
- information bottleneck
- theoretic framework
- jensen shannon divergence
- multi modality
- kullback leibler divergence
- log likelihood
- information theoretic measures
- training data
- decision trees
- support vector
- feature selection
- relative entropy
- entropy measure
- naive bayes
- image registration
- training set
- minimum description length
- text classification
- computational learning theory
- bregman divergences