On the Utility of Incremental Feature Selection for the Classification of Textual Data Streams.
Ioannis KatakisGrigorios TsoumakasIoannis P. VlahavasPublished in: Panhellenic Conference on Informatics (2005)
Keyphrases
- feature selection
- data streams
- classification accuracy
- support vector
- machine learning
- high dimensionality
- text classification
- incremental learning
- support vector machine
- feature extraction
- feature selection algorithms
- classification models
- feature space
- bayes classifier
- feature set
- discriminative features
- feature subset
- pattern recognition
- irrelevant features
- unsupervised learning
- classification method
- information gain
- classification performances
- method for feature selection
- text categorization
- model selection
- feature vectors
- class separability
- ensemble classifier
- image classification
- feature reduction
- small sample
- sensor networks
- supervised learning
- training set
- feature selection and classification
- data stream classification
- anytime classification
- support vector classification
- utility elicitation
- data sets
- informative features
- stream processing
- bayesian classifier
- itemsets
- sliding window
- svm classifier