A Divergence Criterion for Classifier-Independent Feature Selection.
Naoto AbeMineichi KudoJun ToyamaMasaru ShimboPublished in: SSPR/SPR (2000)
Keyphrases
- feature selection
- feature set
- selection criterion
- feature space
- support vector
- support vector machine
- mutual information
- feature selection algorithms
- divergence measure
- multi class
- feature subset
- bayes classifier
- feature ranking
- input features
- text classification
- feature weighting
- classification accuracy
- knn
- information gain
- model selection
- feature selection and classifier
- class separability
- fold cross validation
- high dimensionality
- text categorization
- machine learning
- classification performances
- classification models
- higher classification accuracy
- feature reduction
- selected features
- dependent features
- decision trees
- feature extraction
- discriminative features
- kullback leibler
- training samples
- minimum error
- naive bayes
- class labels
- unsupervised learning