A new approach for textual feature selection based on N-composite isolated labels.
Samir ElloumiPublished in: Nat. Lang. Eng. (2020)
Keyphrases
- feature selection
- text categorization
- feature set
- multi label
- support vector
- high dimensionality
- model selection
- mutual information
- text classification
- multi class
- feature extraction
- training data
- pairwise
- feature space
- dimensionality reduction
- classification accuracy
- irrelevant features
- method for feature selection
- natural language
- multimedia
- class labels
- feature subset
- information gain
- textual information
- machine learning
- supervised learning
- support vector machine
- training set
- training examples
- keywords
- small sample
- visual representations
- unsupervised feature selection
- data mining