Classifier-independent feature selection on the basis of divergence criterion.
Naoto AbeMineichi KudoJun ToyamaMasaru ShimboPublished in: Pattern Anal. Appl. (2006)
Keyphrases
- feature selection
- feature set
- support vector machine
- selection criterion
- feature space
- text categorization
- model selection
- mutual information
- divergence measure
- feature selection algorithms
- dependent features
- class separability
- classification accuracy
- support vector
- bayes classifier
- feature subset
- machine learning
- feature weighting
- selected features
- dimensionality reduction
- feature ranking
- input features
- text classification
- classification performances
- supervised feature selection
- classification models
- high dimensionality
- unsupervised learning
- irrelevant features
- data sets
- feature selection and classifier
- learning algorithm
- feature extraction
- higher classification accuracy
- feature reduction
- naive bayes
- fold cross validation
- information gain
- multi task
- supervised learning
- small sample
- minimum error
- multi class
- ensemble classifier