Twin Boosting: improved feature selection and prediction.
Peter BühlmannTorsten HothornPublished in: Stat. Comput. (2010)
Keyphrases
- feature selection
- prediction accuracy
- select relevant features
- ensemble classifier
- text categorization
- feature space
- text classification
- ensemble classification
- prediction error
- support vector machine
- multi class
- feature set
- prediction model
- classification accuracy
- support vector
- mutual information
- learning algorithm
- stepwise regression
- sequence prediction
- models built
- irrelevant features
- feature weighting
- ensemble learning
- classification models
- machine learning
- combining multiple
- feature extraction
- multi task
- small sample
- ensemble methods
- face recognition
- automatic relevance determination
- high dimensionality
- unsupervised learning