Identifying (Quasi) Equally Informative Subsets in Feature Selection Problems for Classification: A Max-Relevance Min-Redundancy Approach.
Gülsah KarakayaStefano GalelliSelin Damla AhipasaogluRiccardo TaorminaPublished in: IEEE Trans. Cybern. (2016)
Keyphrases
- feature selection
- high dimensionality
- classification accuracy
- classification models
- support vector
- feature space
- feature extraction
- pattern recognition problems
- text classification
- support vector machine
- classification performances
- feature set
- decision trees
- information retrieval
- machine learning
- supervised learning
- pattern recognition
- model selection
- text categorization
- classification algorithm
- dimensionality reduction
- unsupervised learning
- class labels
- microarray
- feature selection algorithms
- irrelevant features
- class separability
- neural network