Differentiable Feature Selection, A Reparameterization Approach.
Jérémie DonàPatrick GallinariPublished in: ECML/PKDD (3) (2021)
Keyphrases
- feature selection
- text categorization
- objective function
- feature selection algorithms
- text classification
- feature set
- naive bayes
- classification accuracy
- feature extraction
- method for feature selection
- unsupervised feature selection
- mutual information
- irrelevant features
- decision trees
- machine learning
- loss function
- feature subset
- feature space
- microarray data
- forward selection
- feature selection and classification
- point sets
- support vector
- multi class
- preprocessing
- gene expression data
- shape analysis
- model selection
- gene selection
- small sample
- training data
- informative features
- redundant features
- knn